CS 5187 — Software Analysis

Monsoon Semester 2026 · 23 July – 12 November 2026 · Dept. of CSE, IIT Guwahati

Schedule

Weeks 1–3: classes Mon / Tue / Wed. From Week 4 (18 Aug): classes Tue / Wed / Thu — Tue 4:00, Wed 3:00, Thu 2:00 PM in Room 5G4. The Thursday slot now carries regular lectures as well as the two quizzes and Industry Talk 2.

Wk	Mon	Tue	Wed	Thu	Notes
1	Jul 27 · Mon-1 L1. Course intro & motivation	Jul 28 · Tue-1 L2. Static / dynamic / hybrid analysis	Jul 29 · Wed-1 L3. Vulnerability assessment & secure coding	—	Topic 1
2	Aug 3 · Mon-2 L4. Soundness, completeness, undecidability	Aug 4 · Tue-2 L5. Precision, recall, F-measure	Aug 5 · Wed-2 L6. Program representation: AST, CFG, IRs	—	Topic 1
3	Aug 10 · Mon-3 L7. Clang tutorial: libtooling & AST analysis	Aug 11 · Tue-3 L8. LLVM IR	Aug 12 · Wed-3 L9. LLVM IR (continued): CFGs & class hierarchy	—	Topic 1 HW 1 out
—	From Week 4 Classes move to Tuesday / Wednesday / Thursday (announced in L9, Wed 12 Aug).
4	—	Aug 18 · Tue-4 L10. Specifications, safety & liveness	Aug 19 · Wed-4 L11. Hoare logic	Aug 20 · Thu-4 L12. Invariants & specification inference	Topic 1 HW 1 due Aug 21
5	—	Aug 25 · Tue-5 L13. Coverage & mutation testing	Aug 26 · Wed-5 No class — holiday	Aug 27 · Thu-5 L14. Introduction to fuzzing	Topic 1 → 2 HW 2 out
6	—	Sep 1 · Tue-6 No class — Institute Day	Sep 2 · Wed-6 L15. Fuzzing (cont.): grammar-based fuzzing & concurrency testing	Sep 3 · Thu-6 L16. Lattices & fixed points; reaching definitions	Topic 2 → 3
7	—	Sep 8 · Tue-7 L17. Available & very busy expressions	Sep 9 · Wed-7 L18. Live variables & DCE; interval analysis & widening	Sep 10 · Thu-7 Quiz 1 2:00–2:55 PM · Topics 1–3	Topic 3 HW 2 due Sep 11 Quiz 1
—	Sep 13–20 Mid-Semester Examinations
8	—	Sep 22 · Tue-8 L19. Industry Talk 1 Room 5G4 · 4:00–4:55 PM	Sep 23 · Wed-8 L20	Sep 24 · Thu-8 L21	HW 3 out Project proposal due Sep 20 Industry Talk 1
9	—	Sep 29 · Tue-9 L22	Sep 30 · Wed-9 L23	Oct 1 · Thu-9 L24
10	—	Oct 6 · Tue-10 L25	Oct 7 · Wed-10 L26	Oct 8 · Thu-10 L27	HW 3 due Oct 9 Progress meeting 1 · Oct 5–9
11	—	Oct 13 · Tue-11 L28	Oct 14 · Wed-11 L29	Oct 15 · Thu-11 L30. Industry Talk 2 Room 5G4 · 2:00–2:55 PM	HW 4 out Industry Talk 2
12	—	—	Oct 21 · Wed-12 L31	Oct 22 · Thu-12 L32	No Tue class this week (institute calendar).
13	—	Oct 27 · Tue-13 L33	Oct 28 · Wed-13 L34	Oct 29 · Thu-13 L35	HW 4 due Oct 30 Progress meeting 2 · Oct 26–30
14	—	Nov 3 · Tue-14 L36. Industry Talk 3 Room 5G4 · 4:00–4:55 PM	Nov 4 · Wed-14 L37	Nov 5 · Thu-14 Quiz 2 2:00–2:55 PM · Topics 4–8	Progress meeting 3 · Nov 2–6 Industry Talk 3 Quiz 2
15	—	Nov 10 · Tue-15 L38	Nov 11 · Wed-15 L39 last teaching day	Nov 12 · Thu-15 Student project presentations	Final presentation · Thu Nov 12 Paper + repo due Fri Nov 13, 23:59 IST
—	Nov 14–21 End-Semester Examinations

Syllabus

#	Topic	Sessions	Count
1	Foundations & Program Representation (incl. Clang / LLVM IR, specifications, Hoare logic & invariants)	L1–L12	12
2	Testing (coverage, mutation, fuzzing, grammar-based, concurrency)	L13–L15	3
3	Dataflow Analysis	L16–L18	3
4	Pointer Analysis	L20–L21	2
5	Constraint-Based Analysis	L22–L25	4
6	Type Systems	L26–L29	4
7	Symbolic Execution	L31–L34	4
8	Automated Test Generation	L35, L37	2
9	LLM-Assisted Software Analysis	L38–L39	2
—	Industry talks (guest sessions)	L19, L30, L36	3
—	Total	L1–L39	39

Logistics

Classroom: Room 5G4, Core 5
Lecture days: Weeks 1–3: Mon / Tue / Wed · Week 4 onwards (from 18 Aug): Tue / Wed / Thu
Lecture times: Mon 5:00–5:55 PM · Tue 4:00–4:55 PM · Wed 3:00–3:55 PM · Thu 2:00–2:55 PM
Thursday slot: Now a regular lecture day. It also carries Quiz 1 (Thu 10 Sep), Industry Talk 2 (Thu 15 Oct), Quiz 2 (Thu 5 Nov) and the final project presentations (Thu 12 Nov).
Office hours: By appointment · email the instructor/TAs
Discussion: Microsoft Teams (CS 5187 channel) · course email list
Handouts & slides: Course Website
Progress meetings: Microsoft Teams (scheduled 1:1 or per group) or offline
Prerequisites: Data structures, C/C++ or Java, basic OS & compilers

Grading

End-semester exam: 30 %
Mid-semester exam: 25 %
H/W assignments: 10 %
Quizzes (2): 10 % (5 % each)
Term project: 20 % (see Project for internal breakdown)
Attendance: 5 %
Total: 100 %

Quizzes (10 %)

Quiz 1: Thu 10 September 2026, 2:00–2:55 PM · covers Topics 1–3 (Foundations, Testing, Dataflow).
Quiz 2: Thu 5 November 2026, 2:00–2:55 PM · covers Topics 4–8 (Pointer, Constraint-based, Types, SymExec, Test generation).

Term Project (20 %)

You'll pick a widely-used open-source target, build or adapt a program-analysis tool from the course, and try to find real bugs that get accepted upstream. Outcome is a short paper in FSE format with a reproducible GitHub artifact.

Implementation & results: 50 % — Does the analyzer find real vulnerabilities and generate successful exploits or fixes?
Full marks on this component require a bug accepted by upstream maintainers.
Progress meetings & final presentation: 25 % — Can you present and explain your work? Evaluated across 3 progress meetings + the final presentation.
Report (FSE format): 25 % — Is the work clearly described and illustrated?

Team & individual accountability

Group size: TBA
Each member must justify their contribution (a signed statement in the report plus commit history in the repo).
At every progress meeting and at the final presentation, each member must be able to answer any question about any part of the project. “That was X's part” is not an acceptable answer.

Timeline

Proposal submitted: Immediately after the mid-semester exam · by Sun 20 Sep 2026 via Microsoft Teams.
Proposal approved: Within one week · do not start work before written approval.
Progress meeting 1: 5 – 9 Oct 2026 · scope firmed up, target OSS accessible, initial pipeline running. Book a slot on Teams.
Progress meeting 2: 26 – 30 Oct 2026 · first end-to-end results, first bug candidates, evaluation plan.
Progress meeting 3: 2 – 6 Nov 2026 · upstream reports/PRs filed, results consolidated, report draft.
Final presentation: Thu 12 Nov 2026 · 10–15 min per group + Q&A · open session on Teams.
Paper + repo due: Fri 13 Nov 2026, 23:59 IST · PDF + public GitHub repo + reproducibility instructions submitted via Teams before end-sem week.

Target open-source project

Must be widely used — think Linux kernel Chromium/V8 FFmpeg OpenSSL SQLite Postgres Node.js CPython LLVM and similar. Popular runtimes, browsers, kernels, media libraries, cryptography libraries, and databases are on-scope.
You cannot pick an obscure repo and simply point your tool at it. Rule of thumb: the project should have a large public user base and a functioning security / bug-triage process.
Get the target on the proposal so I can green-light it before you invest time.

Deliverables

Proposal (2 pages): problem, target OSS project, technique, plan, team & role split. Sent to the instructor by 20 Sep 2026.
Paper in FSE format: motivation, technique, implementation, evaluation, related work, threats to validity. Use the ACM Primary Article Template (LaTeX sigconf). PDF submitted with the artifact.
Public GitHub repo containing:
- Source code with a permissive license
- README.md with build + run instructions, exact dependencies, and one-command reproduction of the headline result
- Docker image or CI workflow that reproduces the evaluation from scratch
- Bug reports / PRs / patches you filed upstream, linked with their state (open / acknowledged / merged)

What earns full marks on the “implementation” component

Your tool finds at least one real, previously-unknown bug in the target project and
The bug report / patch is accepted by upstream maintainers (merged PR, accepted CVE, or acknowledged advisory).
Partial credit is available for high-quality tooling and reproducible evaluation even if no bug lands; the ceiling on this component is capped without an upstream-accepted finding.

Staff

Instructor

Bernard Nongpoh · bnongpoh@iitg.ac.in

Teaching assistants

Shruti Yadav shruti.yadav@iitg.ac.in
Kunal Upadhyay k.upadhyay@iitg.ac.in
Rhythm Jamwal rhythm.jamwal@iitg.ac.in
Yatharth Sonteke y.sonteke@iitg.ac.in
Leisel Rangsuo Koireng leisel.koireng@iitg.ac.in

Policies

Collaboration. Discuss ideas freely; write up assignments individually. Cite any external references you consulted. Academic honesty. Standard institute rules apply. LLM use must be disclosed and cannot replace your own understanding of the material.