Upgrade to Pro
— share decks privately, control downloads, hide ads and more …
Speaker Deck
Features
Speaker Deck
PRO
Sign in
Sign up for free
Search
Search
If You Can't Kill a Supermutant, You Have a Pr...
Search
Rahul Gopinath
April 09, 2018
Research
0
320
If You Can't Kill a Supermutant, You Have a Problem
Mutation Workshop 2018
Rahul Gopinath
April 09, 2018
Tweet
Share
More Decks by Rahul Gopinath
See All by Rahul Gopinath
Assessing Reliability of Statistical Maximum Coverage Estimators in Fuzzing
rahulgopinath
0
2
Fuzzing Without Specifications: Learning Structure from Behaviour Part I
rahulgopinath
0
23
Fuzzing Without Specifications: Learning Structure from Behaviour Part II
rahulgopinath
0
19
How to Compare Fuzzers
rahulgopinath
0
100
Mutation Analysis: Answering the Fuzzing Challenge
rahulgopinath
0
59
An Empirical Evaluation of Frequency Based Statistical Models for Estimating Killable Mutants
rahulgopinath
0
75
Look Ma No Hands! Testing Software Without Specifications
rahulgopinath
0
120
How to Talk to Strange Programs and Find Bugs
rahulgopinath
0
140
Dancing to an Unknown Music: Grammar Inferrence with Prefix Queries
rahulgopinath
0
130
Other Decks in Research
See All in Research
大域マッチングコスト最小化とLiDAR-IMUタイトカップリングに基づく三次元地図生成 / GLIM @ Robotics symposia 2022
koide3
0
100
投資戦略202508
pw
0
560
とあるSREの博士「過程」 / A Certain SRE’s Ph.D. Journey
yuukit
11
4.3k
最適決定木を用いた処方的価格最適化
mickey_kubo
4
1.9k
時系列データに対する解釈可能な 決定木クラスタリング
mickey_kubo
2
1k
J-RAGBench: 日本語RAGにおける Generator評価ベンチマークの構築
koki_itai
0
620
Learning to (Learn at Test Time): RNNs with Expressive Hidden States
kurita
1
250
2025年度人工知能学会全国大会チュートリアル講演「深層基盤モデルの数理」
taiji_suzuki
25
19k
電通総研の生成AI・エージェントの取り組みエンジニアリング業務向けAI活用事例紹介
isidaitc
1
1k
言語モデルの地図:確率分布と情報幾何による類似性の可視化
shimosan
5
1.6k
スキマバイトサービスにおける現場起点でのデザインアプローチ
yoshioshingyouji
0
240
EarthDial: Turning Multi-sensory Earth Observations to Interactive Dialogues
satai
3
210
Featured
See All Featured
Agile that works and the tools we love
rasmusluckow
331
21k
Build your cross-platform service in a week with App Engine
jlugia
232
18k
GitHub's CSS Performance
jonrohan
1032
460k
Navigating Team Friction
lara
189
15k
Building Applications with DynamoDB
mza
96
6.6k
Building Better People: How to give real-time feedback that sticks.
wjessup
368
20k
Rebuilding a faster, lazier Slack
samanthasiow
84
9.2k
Docker and Python
trallard
46
3.6k
JavaScript: Past, Present, and Future - NDC Porto 2020
reverentgeek
52
5.6k
BBQ
matthewcrist
89
9.8k
Let's Do A Bunch of Simple Stuff to Make Websites Faster
chriscoyier
507
140k
Making the Leap to Tech Lead
cromwellryan
135
9.5k
Transcript
9. April 2018 Rahul Gopinath, Björn Mathis and Andreas Zeller
If You Can't Kill a Supermutant You Have a Problem 1
Mutant Kill Matrix 2
Mutant Kill Matrix 2
Mutant Kill Matrix 2
Mutant Kill Matrix 2
Mutant Kill Matrix 2
Mutant Kill Matrix 2
Mutant Kill Matrix 2
Mutant Kill Matrix 2
Mutant Kill Matrix 2 Enormous Amount of Test Executions
Supermutants - Coverage Adequate Test Suite 3
Supermutants - Coverage Adequate Test Suite 0 % 25 %
50 % 75 % 100 % Full Suite (7 Tests) Coverage Adequate 0 % 25 % 50 % 75 % 100 % Full Suite (7 Tests) Coverage Adequate 3
Supermutants - Coverage Adequate Test Suite 0 Coverage Adequate Tests
0 % 25 % 50 % 75 % 100 % Full Suite (7 Tests) Coverage Adequate 0 % 93 % 0 Coverage Adequate Tests 0 % 25 % 50 % 75 % 100 % Full Suite (7 Tests) Coverage Adequate 0 % 93 % 3
Supermutants - Coverage Adequate Test Suite 1 Coverage Adequate Test
0 % 25 % 50 % 75 % 100 % Full Suite (7 Tests) Coverage Adequate 70 % 93 % 1 Coverage Adequate Test 0 % 25 % 50 % 75 % 100 % Full Suite (7 Tests) Coverage Adequate 70 % 93 % 3
Supermutants - Coverage Adequate Test Suite 2 Coverage Adequate Tests
0 % 25 % 50 % 75 % 100 % Full Suite (7 Tests) Coverage Adequate 93 % 93 % 2 Coverage Adequate Tests 0 % 25 % 50 % 75 % 100 % Full Suite (7 Tests) Coverage Adequate 93 % 93 % 3
Supermutants - Coverage Adequate Test Suite 4
Supermutants - Coverage Adequate Test Suite 4
Supermutants - Coverage Adequate Test Suite 4
Supermutants - Coverage Adequate Test Suite 4
Supermutant Execution 5
Supermutant Execution 5 7 Tests
Supermutant Execution 5 7 Tests
Supermutant Execution 5 7 Tests
Supermutant Execution 5 7 Tests
Supermutant Execution 5 3 Tests 3 Tests 7 Tests
Supermutant Execution 5 3 Tests 3 Tests 7 Tests
Supermutant Execution 5 3 Tests 3 Tests 7 Tests
Supermutant Execution 5 3 Tests 3 Tests 7 Tests
Supermutant Execution 5 3 Tests 3 Tests 3 Tests 3
Tests 7 Tests
Supermutant Execution 5 3 Tests 3 Tests 3 Tests 3
Tests 7 Tests
Supermutant Execution 5 3 Tests 3 Tests 3 Tests 3
Tests 7 Tests
Supermutant Execution 5 3 Tests 3 Tests 3 Tests 3
Tests 1 Test 1 Test 7 Tests
Supermutant Execution 5 3 Tests 3 Tests 3 Tests 3
Tests 1 Test 1 Test 7 Tests
Supermutant Execution 5 3 Tests 3 Tests 3 Tests 3
Tests 1 Test 1 Test 7 Tests 34 vs 42 Test Executions
Supermutants - Mutant Kill Matrix 6 7 Tests 3 Tests
3 Tests 3 Tests 3 Tests 1 Test 1 Test
Supermutants - Mutant Kill Matrix 6 Test 1 Test 2
Test 3 Test 4 Test 5 Test 6 Test 7 7 Tests 3 Tests 3 Tests 3 Tests 3 Tests 1 Test 1 Test
Supermutants - Mutant Kill Matrix 6 Test 1 Test 2
Test 3 Test 4 Test 5 Test 6 Test 7 ✕ ✕ ✕ ✕ ✕ ✕ ✕ 7 Tests 3 Tests 3 Tests 3 Tests 3 Tests 1 Test 1 Test
Supermutants - Mutant Kill Matrix 6 Test 1 Test 2
Test 3 Test 4 Test 5 Test 6 Test 7 ✕ ✕ ✕ ✕ ✕ ✕ ✕ ✓ ✕ ✕ ✕ ✕ ✕ ✕ 7 Tests 3 Tests 3 Tests 3 Tests 3 Tests 1 Test 1 Test
Supermutants - Mutant Kill Matrix 6 Test 1 Test 2
Test 3 Test 4 Test 5 Test 6 Test 7 ✕ ✕ ✕ ✕ ✕ ✕ ✕ ✓ ✕ ✕ ✕ ✕ ✕ ✕ ✕ ✕ ✓ ✕ ✕ ✓ ✕ ✕ ✕ ✕ ✕ ✕ ✕ ✕ ✕ ✕ ✓ ✓ ✓ ✓ ✓ ✕ ✕ ✕ ✕ ✕ ✕ ✕ 7 Tests 3 Tests 3 Tests 3 Tests 3 Tests 1 Test 1 Test
Supermutants - Mutant Kill Matrix 6 Test 1 Test 2
Test 3 Test 4 Test 5 Test 6 Test 7 ✕ ✕ ✕ ✕ ✕ ✕ ✕ ✓ ✕ ✕ ✕ ✕ ✕ ✕ ✕ ✕ ✓ ✕ ✕ ✓ ✕ ✕ ✕ ✕ ✕ ✕ ✕ ✕ ✕ ✕ ✓ ✓ ✓ ✓ ✓ ✕ ✕ ✕ ✕ ✕ ✕ ✕ 7 Tests 3 Tests 3 Tests 3 Tests 3 Tests 1 Test 1 Test
Case Study - Urlparse 7
Case Study - Urlparse 7 0 exec. 10000 exec. 20000
exec. 30000 exec. 40000 exec. Full Matrix Strong Testsuite Weak Testsuite
Case Study - Urlparse 7 0 exec. 10000 exec. 20000
exec. 30000 exec. 40000 exec. Full Matrix Strong Testsuite Weak Testsuite 32,000
Case Study - Urlparse 7 0 exec. 10000 exec. 20000
exec. 30000 exec. 40000 exec. Full Matrix Strong Testsuite Weak Testsuite 13,009 32,000
Case Study - Urlparse 7 0 exec. 10000 exec. 20000
exec. 30000 exec. 40000 exec. Full Matrix Strong Testsuite Weak Testsuite 9,037 13,009 32,000
Case Study - Urlparse 7 0 exec. 10000 exec. 20000
exec. 30000 exec. 40000 exec. Full Matrix Strong Testsuite Weak Testsuite 9,037 13,009 32,000 Over 70% of Test Executions Are Saved
When Are Supermutants Super? 8
When Are Supermutants Super? 8 Many and Weak Tests
When Are Supermutants Super? 8 Many and Weak Tests Stubborn
Mutants
When Are Supermutants Super? 8 Many and Weak Tests Stubborn
Mutants No Fault Masking
When Are Supermutants Super? 8 Many and Weak Tests Stubborn
Mutants No Fault Masking Automatically Generated Test Suites Can Be Efficiently Analyzed
9 Summary
9 Summary
9 Summary
9 Summary
9 Summary