Skip to content

QUICKSTART.md

Latest commit

 

History

History
208 lines (152 loc) · 4.35 KB

QUICKSTART.md

File metadata and controls

208 lines (152 loc) · 4.35 KB
Raw

VMM Simulator - Quick Start Guide

Authors: Aditya Pandey, Kartik, Vivek, Gaurang

This guide will get you up and running with the VMM Simulator in 5 minutes!

Option 1: Web GUI (Recommended for Beginners)

Step 1: Navigate to GUI directory

cd gui

Step 2: Start web server

python3 -m http.server 8000

Step 3: Open browser

Navigate to: http://localhost:8000

Step 4: Use the GUI

  1. Configure RAM size and algorithm in the configuration panel
  2. Click "Generate Trace" to create a memory access pattern
  3. Click "Run Simulation" to watch the visualization in real-time
  4. Observe TLB, page table, and frame visualizations update live!

That's it! No compilation needed.

Option 2: Command Line (For Advanced Users)

Step 1: Build the project

make

Step 2: Generate sample traces

make traces

Step 3: Run your first simulation

./bin/vmm -r 64 -p 4096 -t traces/working_set.trace -a LRU -T 64

Step 4: View results

The simulation will output performance metrics to the console:

  • Page faults and fault rate
  • TLB hits/misses
  • Swap I/O statistics
  • Average memory access time

Sample Commands

Small RAM with FIFO

./bin/vmm -r 16 -p 4096 -t traces/sequential.trace -a FIFO -T 16

Expected: High page fault rate with limited memory

Large RAM with LRU

./bin/vmm -r 128 -p 4096 -t traces/locality.trace -a LRU -T 128

Expected: Low fault rate, excellent TLB hit rate

Compare algorithms

# Run with each algorithm
for algo in FIFO LRU CLOCK OPT; do
    ./bin/vmm -r 32 -p 4096 -t traces/random.trace -a $algo -T 32 \
        --csv results_$algo.csv --config-name $algo
done

Result: CSV files for each algorithm that you can compare

Understanding the Output

Console Output

==================== SIMULATION SUMMARY ====================

Memory Accesses:
  Total:              10000
  Reads:               8000 (80.0%)
  Writes:              2000 (20.0%)

Page Faults:
  Total:                 324
  Major:                  42 (required swap-in)
  Minor:                 282 (no I/O)
  Fault Rate:           3.24%   ← Lower is better

TLB Performance:
  Hits:                 9456
  Misses:                544
  Hit Rate:            94.56%   ← Higher is better

Swap I/O:
  Swap-ins:               42
  Swap-outs:              38
  Replacements:          324

Average Memory Access Time:
  AMT:               145.32 ns  ← Lower is better

What These Numbers Mean

  • Fault Rate < 5%: Excellent - your RAM is sufficient

  • Fault Rate 5-20%: Good - reasonable performance

  • Fault Rate > 20%: Poor - thrashing likely occurring

  • TLB Hit Rate > 90%: Excellent locality

  • TLB Hit Rate 70-90%: Good

  • TLB Hit Rate < 70%: Poor locality or TLB too small

Running Tests

Run all automated tests

make test

This runs 10 comprehensive tests including:

  • Algorithm correctness
  • TLB effectiveness
  • Multi-process handling
  • Thrashing scenarios

Next Steps

Learn More

Experiment

  1. Try different trace patterns (sequential, random, locality)
  2. Compare algorithms with the same trace
  3. Vary RAM size and see impact on performance
  4. Test different page sizes (4KB vs 8KB vs 16KB)

Customize

  1. Generate your own traces with ./bin/trace_gen
  2. Create custom trace files (format: pid op addr)
  3. Modify replacement algorithms in src/replacement.c
  4. Add new metrics in src/metrics.c

Troubleshooting

Build fails

# Ensure you have GCC installed
gcc --version

# Clean and rebuild
make clean
make

No traces found

# Generate them
make traces

GUI doesn't work

  • Try a different browser (Chrome recommended)
  • Check if port 8000 is available
  • Use a different port: python3 -m http.server 8080

Getting Help

  1. Check TESTPLAN.md for expected behavior
  2. Read error messages carefully
  3. Enable debug mode: ./bin/vmm ... -D
  4. Run with valgrind: make valgrind

Credits

Developed by: Aditya Pandey, Kartik, Vivek, Gaurang

For educational and research purposes.

Enjoy exploring virtual memory management! 🎉