Chapters

Twenty-five chapters in six parts, from lexing to instruction scheduling. The machine-readable source of this index is tools/course/chapters.yaml, and ./course status shows your progress through it.

Every chapter folder has the same shape (see docs/authoring/STYLE.md §2):

File	What it is
README.md	Technique map: the problem, the family of techniques that solve it, who uses what, and a comparison table. Start here.
lessons/NN-*.md	One lesson per technique family, each with the nine depth-contract sections: algorithm, traced example, correctness, complexity, variants, real compilers, comparison, assessment.
exercises.md	Implementation steps for pebble/ and the comparison lab, with collapsible hints.
flashcards.tsv	Cards for ./course flash NN or Anki (./course flash export).
quiz.yaml	The theory test (hashed answers): ./course quiz NN.

A chapter is done when its quiz score is at least 80 % and all of its implementation tests pass (./course test NN).

All 25 chapters are available. (./course status marks a chapter planned only if its entry in chapters.yaml says so; none does today.)

Part 0 — Orientation

Ch	Chapter	What you learn	Drills
0	Compiler Architectures & Your Toolchain	Pipeline shapes, interpreters vs VMs vs AOT vs JIT, retargeting, and a first tour of the LLVM tools.	phases, stack-code, dispatch, tdiagram

Part I — Front End

Ch	Chapter	What you learn	Drills
1	Lexical Analysis	Regex to NFA to DFA, minimization, derivatives, maximal munch, context-sensitive and lossless lexing.	epsilon-closure, subset-construction, dfa-minimize, brzozowski-derivative, maximal-munch, utf8, lexer-modes
2	Grammars & Top-Down Parsing	CFGs and ambiguity, nullable/FIRST/FOLLOW, LL(1) tables, grammar transformations, recursive descent, ALL(*).	derivations, first-follow, ll1-table, left-recursion, predict-trace, lookahead
3	Bottom-Up Parsing	Shift-reduce, the LR family (LR(0), SLR, LALR, LR(1), IELR), conflicts, GLR and LR error recovery.	lr0-closure, lr-table, shift-reduce-trace, lalr-lookaheads, lr-classify
4	Parsing in Practice: Expressions, Other Paradigms, Recovery & Syntax Trees	Pratt and shunting-yard, PEG/packrat, Earley, resilient and incremental parsing, and syntax-tree design.	pratt-trace, shunting-yard, packrat-memo, earley-chart, cyk-table, paradigm-accepts
5	Names, Scopes & Semantic Analysis	Scoping, symbol-table designs, name-resolution strategies, attribute grammars and definite assignment.	resolve-scopes, attr-eval-order, definite-assignment, edit-distance
6	Type Systems & Type Checking	Typing judgments and soundness, syntax-directed vs bidirectional checking, subtyping, narrowing, generics.	typing-derivation, bidir-modes, subtype-query, progress-preservation, narrowing
7	Type Inference	Unification, Hindley-Milner (Algorithms W, J, M), constraint-based inference, and Swift's solver.	unification, hm-trace, generalization

Part II — Intermediate Representations & LLVM

Ch	Chapter	What you learn	Drills
8	The Design Space of Intermediate Representations	TAC, bytecode, CFG+SSA, block arguments, sea of nodes, CPS/ANF, MLIR; leaders, CFGs and traversal orders.	leaders, rpo, value-numbering, phi-to-block-args, tac-to-anf, stack-code
9	LLVM IR in Depth	Modules, types, SSA values, GEP in full, calls and attributes, metadata, and the verifier's invariants.	gep-offset, ir-validity, poison-propagation, flags
10	The LLVM C++ API	Ownership and intrusive lists, Value/User/Use and RAUW, IRBuilder, LLVM-style RTTI vs C++ RTTI vs variant, PatternMatch vs rule DSLs, ADTs, Error/Expected, interpreter vs MCJIT vs ORC.	use-lists, iterator-invalidation, irbuilder-fold, cast-semantics, pattern-match, densemap-probe, adt-costs, view-lifetime, must-check, jit-compile-set
11	Lowering & Code Generation: AST → PIR → LLVM IR	SSA generation (allocas + mem2reg, on-the-fly), jumping code, switch lowering, places vs values, aggregates and the ABI, safety checks, EH and closures (overview), runtime and linking; native executables.	jumping-code, switch-lowering, abi-classify

Part III — Analysis Foundations & SSA

Ch	Chapter	What you learn	Drills
12	Passes, Pass Managers & Testing Compilers	Pass-manager architectures, plugins, FileCheck/lit, differential testing, reduction, Alive2 and fuzzing.	pm-invalidation, filecheck-match, ddmin-trace, peephole-verify
13	Local Optimization & Transformation Correctness	Folding, peephole engines, canonicalization, LVN, reassociation, and poison/undef/freeze refinement.	lvn-table, rewrite-validity, magic-division, reassoc-ranks
14	Dataflow Analysis & Abstract Interpretation	Lattices and monotone frameworks, worklist and elimination solvers, the four classics, intervals, Datalog.	dataflow-table, worklist-trace, lattice-props, widening
15	Control-Flow Analysis: Dominance & Loops	Iterative, Cooper-Harvey-Kennedy and Lengauer-Tarjan dominators, frontiers, loop forests and reducibility.	dominators, rpo, lengauer-tarjan, dom-update, idf, post-dominance, natural-loops, loop-forms
16	Static Single Assignment Form	Minimal/pruned SSA, Cytron vs Braun construction, phi vs block arguments, destruction and SSA extensions.	phi-placement, ssa-renaming, parallel-copy, lost-copy-swap, idf

Part IV — Optimization

Ch	Chapter	What you learn	Drills
17	SSA-Based Scalar Optimizations	SCCP, ADCE, EarlyCSE and GVN variants, PRE/LCM, SimplifyCFG, and equality saturation.	sccp-trace, vn-partition, lcm-sets, adce-marking
18	Loop Optimizations	LICM, induction variables and SCEV, strength reduction, loop restructuring, dependence analysis, vectorization.	licm-legality, scev-form, trip-count, dependence-test, interchange-legal
19	Memory: Alias Analysis & Memory Optimizations	TBAA and BasicAA, Andersen vs Steensgaard, sensitivity choices, MemorySSA, DSE and load forwarding.	points-to-andersen, points-to-steensgaard, memoryssa-build, tbaa-query
20	Interprocedural & Whole-Program Optimization	Call graphs (CHA/RTA), SCC traversal, inlining cost models, IPSCCP, attribute inference, LTO and PGO.	scc-order, cha-rta, inline-decision, ifds-tabulation, funcattrs-fixpoint

Part V — Back End

Ch	Chapter	What you learn	Drills
21	Instruction Selection & the LLVM Code Generator	Maximal munch vs DP tiling, BURS, SelectionDAG vs GlobalISel, ISLE, TableGen and the MC layer.	munch-tiling, dp-tiling, legalization, calling-convention
22	Register Allocation	Local, Chaitin-Briggs/IRC, linear scan, SSA-based and PBQP allocation, and LLVM's greedy allocator.	interference-graph, chaitin-briggs, coalescing-test, linear-scan, belady, ssa-coloring, pbqp
23	Instruction Scheduling & Machine-Level Optimization	Dependence DAGs, list scheduling, trace/superblock scheduling, modulo scheduling, MachineScheduler, BOLT.	critical-path, list-schedule, trace-select, ifconvert, mii, modulo-table, code-layout

Part VI — Capstone

Ch	Chapter	What you learn	Drills
24	The Complete Pebble Compiler — and Beyond	Your own -O1 pipeline end to end, differential fuzzing, JIT designs, debug info, memory management.	pass-order, gc-roots, dwarf-location, refcount-trace

Suggested paths

• Full course: chapters 0 → 24 in order.
• IR/optimization fast track: 0 → 8 → 9 → 10 → 12 → 13 → 14 → 15 → 16 → 17 → 18 → (19, 20) → 24. Build the front end from solutions/ (-DPEBBLE_USE_SOLUTION=…, see the course home) and come back to 1–7 and 11 later.