Correct answer: A. Translate program representations

A compiler translates a program from a source representation to another representation while preserving the intended semantics of valid programs. The output may be machine code, bytecode, or an intermediate language.

Correct answer: B. Lexical syntax and semantic analysis

Front-end phases typically scan source into tokens, check grammatical structure, and perform semantic analysis such as name resolution and type checking. They build a representation for later compilation.

Correct answer: C. Target optimization and code generation

A back end converts an intermediate representation into target-specific code. It commonly performs instruction selection, register allocation, scheduling, and target-specific optimizations.

Correct answer: D. Executes program representations

An interpreter directly evaluates source or an intermediate representation rather than first producing a separate ahead-of-time executable. Some interpreters combine interpretation with just-in-time compilation.

Correct answer: A. Targets another platform

A cross-compiler runs on a host platform but produces code for a different target platform. For example, a desktop compiler may emit instructions for an embedded processor.

Correct answer: B. Token sequence

Lexical analysis groups source characters into tokens such as identifiers, keywords, literals, and operators. The parser then works with this token stream rather than raw characters.

Correct answer: C. Class of lexeme

A token category identifies a class of lexemes, such as IDENTIFIER or NUMBER. The lexeme is the particular source text that matched the category.

Correct answer: D. Matched source substring

A lexeme is the exact sequence of source characters recognized for one token. In `count + 1`, the substring `count` is an identifier lexeme.

Correct answer: A. Name for programmer-defined entity

Identifiers are names assigned to program entities such as variables, functions, or types. Which character sequences are legal depends on the language’s lexical rules.

Correct answer: B. Language-defined grammatical word

A keyword is a word with a language-defined syntactic role, such as `while` or `return` in many languages. Reserved keywords cannot be used freely as identifiers.

Correct answer: C. Describe lexical patterns

Regular expressions describe patterns for token classes. Lexer generators commonly translate those patterns into finite automata for efficient scanning.

Correct answer: D. Recognize regular languages

A DFA recognizes precisely the regular languages. Its next state is determined by the current state and the next input symbol.

Correct answer: A. Subset construction

Subset construction represents sets of NFA states as DFA states. The resulting deterministic automaton accepts the same regular language as the original NFA.

Correct answer: B. Longest valid token prefix

Maximal munch chooses the longest prefix of the remaining input that matches a token rule. This typically avoids splitting a longer valid operator into shorter tokens.

Correct answer: C. Resolves equal-length matches

When two token rules match the same longest lexeme, an explicit priority rule determines which category wins. This is useful for keyword-versus-identifier ties.

Correct answer: D. Skip or tokenize by language rules

Whitespace may be ignored, emitted as tokens, or used to delimit structure depending on the language. Indentation-sensitive languages require special treatment.

Correct answer: A. Unmatched character sequence

A lexical error occurs when the scanner cannot match the next input characters to any valid token rule. A lexer may report the location and attempt recovery.

Correct answer: B. Marks logical input end

An EOF token marks the end of the token stream. Parsers may use it to confirm that a complete grammar start symbol has been recognized without extra input.

Correct answer: C. Records source text position

Source spans record positions such as line, column, or character offsets for a lexeme or syntax node. Diagnostics use them to point users to the source of an error.

Correct answer: D. Generates lexical scanners

Flex generates a lexical scanner from rules usually written as regular-expression patterns paired with actions. The generated scanner recognizes tokens in input text.

Correct answer: A. Tokens

A parser usually consumes a token sequence produced by the lexer. Token categories and attached values allow it to check syntax and build structured representations.

Correct answer: B. Productions over terminals and nonterminals

A CFG contains terminal and nonterminal symbols, production rules, and a start symbol. It specifies which strings have syntactically valid derivations.

Correct answer: C. Shows grammar derivation structure

A parse tree records how grammar productions derive the input. Its internal nodes correspond to nonterminals, while its leaves yield the terminal sequence.

Correct answer: D. Grammar start symbol

The root of a complete parse tree is the grammar’s start symbol because the entire input is derived from that symbol.

Correct answer: A. Terminal leaves in order

The yield is the sequence of terminal leaves read left to right. It spells the generated input sentence, ignoring epsilon leaves if present.

Correct answer: B. One string admits two parse trees

A grammar is ambiguous if at least one terminal string has two distinct parse trees. Different structures can imply different meanings for the same expression.

Correct answer: C. Replace leftmost nonterminal

In a leftmost derivation, each step expands the leftmost remaining nonterminal. The choice of productions determines the final sentence.

Correct answer: D. Replace rightmost nonterminal

In a rightmost derivation, the rightmost remaining nonterminal is expanded at each step. LR parsing reconstructs such a derivation in reverse.

Correct answer: A. Omits unnecessary syntax nodes

An AST usually omits punctuation and grammar-specific scaffolding that do not affect program structure. For example, nested expression nodes retain the operator hierarchy.

Correct answer: B. Token sequence violates grammar

A syntax error is a token sequence that cannot be derived from the grammar. Examples include unmatched delimiters or an operator appearing where an operand is required.

Correct answer: C. Expands from start symbol

Top-down parsing begins with the start symbol and expands nonterminals to match the input. It attempts to construct a leftmost derivation.

Correct answer: D. Reduces input toward start symbol

Bottom-up parsing starts from the input tokens and repeatedly reduces recognized handles toward the grammar’s start symbol. This reverses a rightmost derivation.

Correct answer: A. Grammar-directed recursive procedures

Recursive-descent parsers implement nonterminals as mutually recursive procedures. The procedures match terminals and call each other according to grammar rules.

Correct answer: B. Uses lookahead to select production

A predictive parser uses the current nonterminal and lookahead token to choose a production without arbitrary backtracking when the grammar supports prediction.

Correct answer: C. Left-to-right leftmost with one lookahead

LL(1) means reading input Left to right, producing a Leftmost derivation, with one lookahead token. A suitable grammar permits an unambiguous table choice.

Correct answer: D. Possible beginning terminals

FIRST describes terminals that may begin strings derivable from a symbol or sequence. It can also contain epsilon if the sequence may derive the empty string.

Correct answer: A. Possible terminals after a nonterminal

FOLLOW(A) is the set of terminals that may occur immediately after nonterminal A in some sentential form. It may include the end marker for the start symbol.

Correct answer: B. Can derive epsilon

A nonterminal is nullable when it can derive the empty string epsilon. Nullable symbols influence FIRST and FOLLOW computations.

Correct answer: C. Naive recursive descent may not consume input

A production like A → Aα can call the same recursive-descent procedure before consuming input. Without transformation or special handling, this may recurse indefinitely.

Correct answer: D. Extracts common prefixes

Left factoring extracts a shared prefix from competing productions. The parser can postpone the choice until enough input distinguishes the alternatives.

Correct answer: A. Nonterminal and lookahead token

A table-driven predictive parser indexes its decision using the current nonterminal and lookahead terminal. The selected cell identifies the production to apply.

Correct answer: B. Two choices compete in one table cell

An LL(1) conflict occurs when a table cell requires more than one production. The grammar therefore cannot be parsed deterministically by that LL(1) table.

Correct answer: C. Start symbol and end marker

An LL parser typically initializes its stack with the grammar start symbol and an end-of-input marker. Matching proceeds by expansion or terminal consumption.

Correct answer: D. Discard or synchronize tokens

Panic-mode recovery discards input tokens or pops grammar symbols until a synchronization point is reached. It helps produce subsequent diagnostics after an error.

Correct answer: A. Lexer groups characters; parser checks token syntax

The lexer recognizes words or token patterns from characters; the parser checks how those tokens combine according to grammar productions.

Correct answer: B. Additional value such as numeric literal value

Token attributes hold information beyond a category, such as the numeric value of an integer literal or an entry for an identifier in a symbol table.

Correct answer: C. Represents progress in token pattern recognition

A lexer state indicates how far scanning has progressed through possible token patterns. Accepting states identify successful token matches.

Correct answer: D. Upcoming token used to choose action

Lookahead examines upcoming tokens before committing to an action. It helps select productions in predictive parsing or shift/reduce decisions in LR parsing.

Correct answer: A. An atomic token in grammar productions

A terminal symbol is an atomic symbol of the grammar that corresponds to a token consumed from the input. It is not expanded by productions.

Correct answer: B. A syntactic category expanded by productions

A nonterminal represents a syntactic category such as Expression or Statement. Its production rules specify how it may expand.

Correct answer: C. Shifts tokens and reduces handles

A shift-reduce parser repeatedly shifts input tokens onto a stack and reduces recognized handles using grammar productions. These operations construct a bottom-up parse.

Correct answer: D. Scans left to right and constructs rightmost derivation in reverse

LR parsing reads tokens from left to right and reconstructs a rightmost derivation in reverse. LR tables can handle many grammars that are difficult for LL parsers.

Correct answer: A. Production with a dot indicating parse progress

An LR(0) item is a production with a dot marking how much of its right-hand side has been recognized. For example, A → a·b indicates that a has been seen.

Correct answer: B. Adds items for nonterminals following dots

Closure adds items for productions of nonterminals immediately after dots, repeating until no further items can be added. It captures possible parser continuations.

Correct answer: C. Moves item sets on a grammar symbol

GOTO(I, X) advances dots over grammar symbol X in the relevant items of state I and then computes closure. The result is a successor parser state.

Correct answer: D. Chooses shift reduce accept or error

An LR ACTION table uses the current state and lookahead terminal to choose shift, reduce, accept, or error. Its conflicts reveal unresolved parsing choices.

Correct answer: A. Chooses next state after reducing a nonterminal

After reducing a right-hand side to nonterminal A, the parser consults GOTO using the exposed stack state and A to choose the next state.

Correct answer: B. State and lookahead permit both shift and reduction

A shift-reduce conflict occurs when one state and lookahead allow both shifting and reducing. Precedence rules or grammar changes may resolve it.

Correct answer: C. Two reductions compete at a parsing point

A reduce-reduce conflict occurs when multiple grammar productions could be reduced in the same state with the same lookahead. It creates an ambiguous parser decision.

Correct answer: D. Uses LR(0) items and FOLLOW sets for reductions

SLR uses LR(0) item states and a completed item’s left-hand-side FOLLOW set to authorize reductions. The approximation may introduce conflicts.

Correct answer: A. Contains a production position and one-token lookahead

An LR(1) item includes a lookahead terminal in addition to the production and dot position. This more precise context can avoid some SLR conflicts.

Correct answer: B. Merges LR(1) states with identical LR(0) cores

LALR typically merges canonical LR(1) states sharing the same LR(0) core, combining their lookahead sets. This reduces table size but can introduce conflicts.

Correct answer: C. Can resolve certain expression grammar conflicts

Precedence declarations guide choices such as shift versus reduce for expression operators. They may remove conflicts without rewriting the grammar.

Correct answer: D. Conventionally left associative in many languages

In many languages subtraction is left associative, so a − b − c is parsed as (a − b) − c. Language specifications determine the actual rule.

Correct answer: A. Else may associate with more than one unmatched if

Nested if statements without explicit delimiters may allow an else to match different preceding if statements. Many languages pair else with the nearest unmatched if.

Correct answer: B. Substring matching RHS of the next reverse-derivation reduction

A handle is the substring that corresponds to the next production right-hand side in the reverse of a rightmost derivation. Reducing it moves parsing toward the start symbol.

Correct answer: C. Prefix that can appear on an LR parser stack in a valid parse

A viable prefix is a prefix of a right-sentential form that does not extend beyond a handle’s end. The LR automaton recognizes such stack prefixes.

Correct answer: D. Generates parser code from grammar specifications

Bison generates parser source code from grammar and action specifications. It commonly uses LR-family parsing strategies and produces diagnostic information.

Correct answer: A. A grammar rule supporting recovery or diagnostics

An error production explicitly describes malformed input patterns so the parser can recover or report a targeted diagnostic. It must be designed carefully to avoid masking errors.

Correct answer: B. Traverses tree nodes to perform operations

An AST visitor organizes operations that walk tree nodes, such as type checking or code generation. It separates traversal behavior from node definitions.

Correct answer: C. Checks program meaning such as types and scope

Semantic analysis checks constraints not captured by grammar alone, including name binding, types, declarations, and some use-context rules.

Correct answer: D. Maps names to attributes such as type and scope

A symbol table associates names with properties such as their declared type, source location, storage class, and scope. It helps resolve identifier references.

Correct answer: A. Region in which a declaration or binding is visible

Scope defines the part of a program in which a binding is visible or eligible for reference. Blocks, functions, and modules may introduce scopes.

Correct answer: B. Resolves names from syntactically enclosing scopes

Lexical scoping resolves a name using scopes determined by the written nesting of the program. It does not depend on which function happened to call another.

Correct answer: C. Resolves bindings according to active runtime call environments

Dynamic scoping looks for an active binding through the runtime call chain. The binding can therefore depend on the sequence of function calls.

Correct answer: D. Inner declaration hides an outer binding with the same name

Shadowing occurs when an inner declaration has the same name as one in an enclosing scope. References in the inner scope generally resolve to the nearer declaration.

Correct answer: A. Determines declaration associated with a use

Name resolution identifies which declaration a use of a name refers to. It may need to account for imports, overloads, visibility and nested scopes.

Correct answer: B. Checks type constraints without ordinary program execution

Static type checking validates type constraints before ordinary execution, often during compilation. It can detect invalid operations without running the program.

Correct answer: C. Checks type requirements during execution

Dynamic type checking validates type requirements during execution using runtime values and metadata. A type error may occur only on a particular execution path.

Correct answer: D. Derives types from expressions and constraints

Type inference derives types from expressions and constraints when annotations are incomplete or omitted. The rules depend on the language’s type system.

Correct answer: A. Determines whether types can be used together under language rules

Type compatibility determines whether values or expressions of certain types can be used together, such as assignment to a variable or passing an argument.

Correct answer: B. Implicit or explicit conversion between types

Coercion converts a value between types, sometimes implicitly. For example, numeric rules may convert an integer to a floating-point representation.

Correct answer: C. Expression denoting a storage location in relevant language models

An lvalue generally identifies a location or object that can be referred to, and sometimes assigned to, depending on the language. It differs from a temporary computed value.

Correct answer: D. Expression value used in computation

An rvalue commonly denotes a value used in an expression rather than an assignable storage location. Exact value-category rules vary by language.

Correct answer: A. Evaluates constant expressions during compilation

Constant folding evaluates compile-time-known expressions, for example replacing 3 × 4 with 12. This saves runtime work when the transformation respects language semantics.

Correct answer: B. Requires name declaration prior to permitted uses in languages enforcing it

Some languages require declarations before uses so the compiler can resolve the name and its type. Other languages permit forward references under defined rules.

Correct answer: C. Each SSA variable name is assigned exactly once

In SSA, each named variable version is assigned exactly once. Creating new versions simplifies analyses of definitions and data flow.

Correct answer: D. Selects SSA value based on incoming control-flow edge

Phi functions choose among versions of a variable according to the incoming control-flow edge at a merge point. This reconciles alternative definitions in SSA.

Correct answer: A. Associates semantic information and rules with grammar symbols

Attribute grammars attach semantic attributes and evaluation rules to grammar symbols. They formalize syntax-directed calculations such as type or value propagation.

Correct answer: B. Computed from child attributes and local information

Synthesized attributes are calculated using child attributes and local information at a parse-tree node. Information generally flows upward.

Correct answer: C. Computed from parent or sibling context under attribution rules

Inherited attributes may use information from a parent or siblings, subject to the grammar’s evaluation rules. They pass contextual information downward or sideways.

Correct answer: D. Uses grammar productions with associated semantic actions

Syntax-directed translation attaches semantic actions to productions. As parsing recognizes constructs, these actions build IR or compute semantic attributes.

Correct answer: A. Compiler-internal form between source and target code

An intermediate representation is a compiler-internal program form between source analysis and target code. It allows reusable optimizations independent of source syntax.

Correct answer: B. Instruction-like IR with limited operands and destinations

Three-address code decomposes complex expressions into simple operations with explicit temporaries, often with at most two input operands and one result.

Correct answer: C. Often stores operator argument1 argument2 result

A quadruple conventionally stores operator, first argument, second argument, and result. It can encode simple three-address instructions for later passes.

Correct answer: D. Straight-line instruction region with single entry and exit in a control-flow sense

A basic block is a maximal straight-line sequence with one entry at its beginning and control leaving at its end, apart from exceptional control mechanisms.

Correct answer: A. Represents basic blocks and possible transfer edges

A CFG has basic blocks as nodes and possible control transfers as edges. It supports reasoning about reachability, loops and program analyses.

Correct answer: B. Node A dominates B if every entry-to-B path passes through A

Node A dominates node B if every path from the CFG entry to B includes A. Dominance is central to SSA construction and optimization.

Correct answer: C. Computes facts across program control-flow paths

Data-flow analysis propagates facts through a CFG with transfer and merge rules. Examples include liveness and reaching-definitions analysis.

Correct answer: D. Mapping from in-scope names to type information

A type environment maps currently accessible names to type information. The type checker consults it when validating expressions and statements.

Correct answer: A. Separates source analysis from target-specific generation

An IR separates language-specific source analysis from target-specific code generation. The same middle-end passes can optimize several source languages or hardware targets.

Correct answer: B. Represents simple computations with explicit temporaries

Three-address code splits compound expressions into simple operations with temporary variables. Explicit intermediate results are easier for analysis and code generation.

Correct answer: C. Represents a possible transfer between basic blocks

A CFG edge indicates that execution can transfer from one basic block to another. Conditional branches can create multiple outgoing edges.

Correct answer: D. Transfers control to a subsequent block without an explicit taken branch

A fall-through edge connects a block to a following block when execution continues without taking an explicit branch. Its meaning depends on emitted code layout.

Correct answer: A. Cannot be reached from entry by any CFG path

An unreachable block has no path from the entry in the CFG. Removing it can simplify subsequent optimization when exceptional control flow is properly modeled.

Correct answer: B. Block that typically dominates nodes in a natural loop

A natural loop header is the entry point for the loop and dominates its nodes. Back edges into the header characterize common reducible loops.

Correct answer: C. Control-flow edge to a node dominating its source

An edge from node X to node H is a back edge when H dominates X. Such edges can help identify loops in a reducible CFG.

Correct answer: D. Organizes control-flow dominance relationships

A dominator tree represents immediate-dominator relationships among reachable blocks. Each parent dominates its children, making dominance queries efficient.

Correct answer: A. Closest strict dominator of a node

The immediate dominator is a node’s closest strict dominator. Every reachable nonentry block has one under the standard single-entry CFG model.

Correct answer: B. Node on every path from a given node to exit under a defined CFG model

A postdominator is a block occurring on every relevant path from another block to the exit. Postdominance is often used for control-dependence analysis.

Correct answer: C. Describe facts created and invalidated by blocks

GEN identifies facts produced by a block, while KILL identifies facts the block invalidates. They form transfer functions in classical bit-vector analyses.

Correct answer: D. Determines assignments that may reach each program point

Reaching-definitions analysis finds assignments whose values might reach a program location without an intervening assignment to the same variable.

Correct answer: A. Determines values potentially used before being overwritten

A variable is live if its current value may be used later before being overwritten. Liveness information helps determine overlapping register lifetimes.

Correct answer: B. Propagates information from successors toward predecessors

Backward analyses propagate facts from successors to predecessors. Live-variable analysis works backward because later uses determine earlier liveness.

Correct answer: C. Propagates facts from predecessors toward successors

Forward analyses propagate information from predecessors to successors. Reaching-definitions analysis is a standard forward-flow example.

Correct answer: D. Orders abstract information by precision or inclusion under a framework

A lattice orders abstract program facts and supplies merge operators such as join or meet. Its properties help establish existence and convergence of iterative solutions.

Correct answer: A. Repeats transfer updates until analysis facts stabilize

Fixed-point iteration repeatedly applies transfer and merge operations until no facts change. At that point the analysis has reached a stable solution.

Correct answer: B. Replaces uses with values proven constant

Constant propagation discovers variables whose values are provably constant at a point and substitutes those values when safe. It can expose more folding opportunities.

Correct answer: C. Replaces a copied variable use with its source when safe

Copy propagation replaces uses of a copied variable by its source where definitions and aliases permit. For example, after x = y, a safe use of x may use y.

Correct answer: D. Removes computations without observable effect under compiler assumptions

Dead-code elimination removes operations that have no required observable effects and whose results are not used. Side effects and exceptions must still be respected.

Correct answer: A. Avoids repeating an equivalent computation when operands remain valid

Common-subexpression elimination reuses a computed value when an equivalent expression has already been evaluated and its operands remain valid.

Correct answer: B. Expressions already computed on all incoming paths without invalidation

An expression is available if it has been computed on every incoming path and none of its relevant operands have subsequently been redefined.

Correct answer: C. Moves or inserts computations to eliminate partially redundant evaluations

Partial-redundancy elimination can move or introduce computations to make repeated evaluations fully redundant, then remove them. Correctness depends on control flow and side effects.

Correct answer: D. Moves safe unchanged computations outside a loop

Loop-invariant code motion moves expressions whose values stay constant across iterations outside the loop when moving them cannot change observable behavior.

Correct answer: A. Substitutes cheaper equivalent operations under valid semantics

Strength reduction replaces an operation with a potentially cheaper equivalent one under applicable semantics, such as simplifying address arithmetic in a loop.

Correct answer: B. Variable changing predictably across loop iterations

An induction variable changes according to a predictable recurrence across loop iterations, often by adding a constant. Such variables enable simpler address calculations.

Correct answer: C. Duplicates iteration bodies to reduce overhead or expose parallelism

Unrolling duplicates a loop body across several iterations. It may reduce branch overhead or expose parallelism at the expense of code size.

Correct answer: D. Swaps nested loop order when dependences permit

Interchanging nested loops swaps their order only when loop-carried dependences and language semantics permit. It can improve access locality.

Correct answer: A. Combines compatible adjacent loops over related iteration spaces

Loop fusion combines compatible loops to reuse data and reduce loop overhead. It is legal only when changing the interleaving of iterations preserves dependencies.

Correct answer: B. Splits a loop into multiple loops when legal

Loop fission divides one loop into multiple loops. This may isolate operations for vectorization or improve cache behavior, but may increase passes over data.

Correct answer: C. Replaces a call with the callee body when valid

Inlining substitutes the body of a called function at its call site. It can eliminate call overhead and enable optimization across former function boundaries.

Correct answer: D. Replaces an indirect virtual dispatch with a known direct target when provable

Devirtualization replaces an indirect virtual call with a direct target when the compiler can prove which method is invoked, sometimes with guards and fallbacks.

Correct answer: A. Examines small instruction windows for local improvements

Peephole optimization examines short sequences of instructions and replaces them with simpler, equivalent sequences. It is typically a local transformation.

Correct answer: B. Can replace eligible final call with frame reuse or jump

An eligible tail call can become a jump or reuse the current stack frame because no further caller computation is needed after the callee returns.

Correct answer: C. Determines whether objects or references outlive or escape their context

Escape analysis checks whether an object or reference can become accessible beyond a local context. Non-escaping objects may allow stack allocation or scalar replacement.

Correct answer: D. Estimates whether references may identify the same storage

Alias analysis determines whether two references can access the same memory location. It helps prove when memory operations may safely move or be eliminated.

Correct answer: A. Two references might point to the same object or location

May-alias means two references might refer to the same location in some execution. The compiler must conservatively account for that possibility.

Correct answer: B. References necessarily designate the same location under analyzed conditions

Must-alias means references definitely denote the same location under the analysis assumptions. This stronger guarantee may support additional transformations.

Correct answer: C. Identifies observable changes or interactions of operations

Side-effect analysis identifies observable consequences such as memory writes, I/O, or exceptions. Optimizations must not remove or reorder required effects.

Correct answer: D. May improve speed but increase code size

Inlining can speed hot call sites, but replicating function bodies increases code size. Excessive size may hurt instruction-cache locality and compilation time.

Correct answer: A. Must preserve required observable program behavior

Optimization must preserve all behavior required by the source language and compilation assumptions. A faster result is invalid if it changes specified observable semantics.

Correct answer: B. Can permit optimizers to assume prohibited executions do not occur under that language model

If the language leaves some operations undefined, an optimizer may assume valid programs do not perform them. This requires care because programmers may expect unintended behavior.

Correct answer: C. Depends on signedness and language-specific overflow rules

Integer overflow rules differ across languages and signed or unsigned types. Algebraic rewrites are unsafe if they change wraparound, trapping, or undefined behavior.

Correct answer: D. May alter rounding and therefore require language or flag permissions

Floating-point calculations round after intermediate operations, so (a+b)+c need not equal a+(b+c). Reassociation may change numerical results unless semantics permit it.

Correct answer: A. Uses measurements from representative runs to guide transformations

Profile-guided optimization uses measurements from representative program runs, such as branch frequencies or hot paths. Unrepresentative profiles can limit benefits.

Correct answer: B. Generates executable code during program execution

JIT compilation generates machine code during program execution. It can use runtime information to specialize and optimize frequently executed code.

Correct answer: C. Generates code ahead of program execution

AOT compilation produces executable or object code before the program runs. It reduces reliance on runtime code generation and may improve startup behavior.

Correct answer: D. Uses information across module boundaries where available

Whole-program optimization uses information across available modules or translation units. Broader visibility can enable cross-module inlining and analysis.

Correct answer: A. Runs compiler optimizations across code at linking stage

Link-time optimization runs compiler analyses or transformations while linking modules, often using stored IR or summaries to make cross-module decisions.

Correct answer: B. Can affect opportunities and resulting performance

Passes can enable or disable opportunities for other passes. For example, inlining may expose constants for propagation, so order can affect code and performance.

Correct answer: C. Maps IR operations into target machine instructions

Code generation lowers intermediate operations to instructions for a target architecture. The result may be assembly, machine code, or a lower-level representation for a later assembler.

Correct answer: D. Chooses target instructions for intermediate operations

Instruction selection chooses machine instructions that implement IR operations while honoring target constraints. Different choices can affect code size and speed.

Correct answer: A. Matches expression trees against target instruction patterns

Tree-pattern selection matches IR expression trees against supported target instruction patterns. A cost model may help choose a low-cost covering.

Correct answer: B. Orders machine operations respecting dependencies and resource constraints

Instruction scheduling reorders independent operations while respecting data dependencies and machine constraints. It aims to reduce stalls and improve pipeline usage.

Correct answer: C. Assigns values to a limited set of machine registers

Register allocation maps temporary values or live ranges to the limited set of hardware registers. Values that do not fit may need to be stored in memory.

Correct answer: D. Moves selected live values to memory when registers are insufficient

Spilling moves selected values to stack or other memory locations because too many values are live for available registers. Loading and storing spills adds overhead.

Correct answer: A. Connects values whose live ranges cannot share one register

An interference graph connects values whose live ranges overlap so they cannot occupy the same register. Each assigned register corresponds to a graph color.

Correct answer: B. Assigns register colors to noninterfering live ranges

Graph coloring assigns distinct register colors to interfering live ranges. If the graph cannot be colored using available registers, the allocator may introduce spills.

Correct answer: C. Combines compatible copy-related live ranges to remove moves

Coalescing combines compatible live ranges connected by copy instructions. Removing such copies is beneficial if the merge does not create unacceptable interference.

Correct answer: D. Allocates registers by processing intervals in program order

Linear scan processes live intervals in an order related to their program positions. It is relatively fast and is often selected by runtime compilers.

Correct answer: A. Defines argument passing returns and saved registers

A calling convention defines how arguments and return values are passed and which registers must survive calls. It is part of the platform’s application binary interface.

Correct answer: B. Must be preserved by a caller when its value is needed across a call

Caller-saved registers may be overwritten by the called function. The caller must save their values when those values are still needed after the call.

Correct answer: C. Must be restored by the called routine if it modifies the register

A callee-saved register must be restored by the called function if it modifies that register. This lets callers rely on specified register values remaining intact.

Correct answer: D. Holds activation-specific data such as locals and saved state

A stack frame commonly contains local variables, saved registers, return-related data, and temporary storage for one invocation. Its exact layout depends on the ABI.

Correct answer: A. Tracks a current stack position or boundary

The stack pointer tracks the current stack boundary or position. Instructions use it to allocate and access stack storage for calls and temporary data.

Correct answer: B. Provides a stable reference point for a call frame in implementations using it

A frame pointer provides a stable base for addressing activation-record contents when stack-pointer changes would complicate offsets. Some optimizers omit it.

Correct answer: C. Establishes necessary stack frame and callee state

The prologue establishes the function’s required calling state, for example adjusting the stack pointer and saving callee-preserved registers.

Correct answer: D. Restores saved state and prepares to return

The epilogue restores saved state, releases the frame as needed, and transfers control back to the caller. Its details follow the ABI.

Correct answer: A. A call in a position where caller has no further computation after it

A tail call appears where the caller has no further computation after the callee returns. Some implementations can reuse the current frame.

Correct answer: B. Supplies a value independent of reassignment to the callee’s parameter binding

Pass-by-value supplies the callee with a value for its parameter binding. Assigning a new value to that binding does not itself reassign the caller’s variable.

Correct answer: C. Allows a parameter to refer to a caller-supplied location or object according to language rules

Pass-by-reference gives the callee access to a caller-associated location or reference according to language rules. Modifications may therefore be visible to the caller.

Correct answer: D. Runtime storage for one procedure invocation

An activation record stores information for one particular procedure invocation. Multiple recursive calls have distinct activation records.

Correct answer: A. Provides access to a lexically enclosing activation environment

Static links connect a nested routine’s activation to an enclosing lexical environment. They make outer-scope variables accessible at runtime.

Correct answer: B. Combines executable function behavior with captured lexical environment

A closure packages executable function behavior with captured bindings from its lexical environment. It can retain access to data after the enclosing call returns.

Correct answer: C. Manages objects whose lifetimes need not follow function calls in stack order

Heap allocation accommodates objects whose lifetimes do not follow the last-in, first-out order of procedure calls. It normally requires explicit or automatic memory management.

Correct answer: D. Often supports last-in-first-out lifetime of activation data

Stack allocation is effective for activation data whose lifetimes follow nested calls. Storage can be reclaimed efficiently as stack frames return.

Correct answer: A. Reclaims unreachable allocated objects under a collector’s reachability model

Garbage collection automatically reclaims memory judged unreachable or otherwise collectable according to the runtime’s memory-management model.

Correct answer: B. Tracks numbers of owning references to objects

Reference counting maintains counts of owning references. A zero count often permits reclamation, although cycles can remain without additional handling.

Correct answer: C. Finds reachable objects from roots to identify garbage

A tracing collector begins from roots such as stacks and globals, follows references to mark reachable objects, and treats unmarked objects as garbage.

Correct answer: D. Separates objects by age or survival characteristics to improve collection efficiency

Generational collectors exploit the observation that many newly allocated objects become unreachable quickly. Young and old regions can be collected at different frequencies.

Correct answer: A. Preserves type-related metadata for runtime operations

RTTI is metadata available at runtime for operations such as dynamic type checking, reflection, or casts. The exact facilities depend on the language and runtime.

Correct answer: B. Describe handler regions and unwind information under relevant runtime formats

Exception tables describe protected code ranges, handlers and unwinding actions. The runtime can consult this metadata when an exception propagates.

Correct answer: C. Contains compiled code data and linking metadata

Object files may contain machine instructions, initialized data, symbol definitions, unresolved references, and relocation entries. A linker combines these into an output artifact.

Correct answer: D. Associates references with appropriate symbol definitions

Symbol resolution matches a reference in one module to an appropriate definition in another or the same module. Visibility and linkage rules govern which definitions qualify.

Correct answer: A. Adjusts address-dependent code or data references after placement

Relocation adjusts encoded references to account for the actual positions of code and data. A linker or loader may apply relocations after layout is chosen.

Correct answer: B. Combines required code into an output artifact at build time

Static linking incorporates the required object code and libraries into the produced artifact at build time. This can reduce reliance on separate runtime libraries.

Correct answer: C. Resolves or loads shared library dependencies through runtime or loader mechanisms

Dynamic linking connects a program to shared library code through loader or runtime mechanisms. Libraries may be updated or shared without embedding every routine in an executable.

Correct answer: D. Uses addressing techniques that allow execution from different load addresses

Position-independent code uses addressing techniques that permit execution at different load addresses. It is useful for shared libraries and address-space layout randomization.

Correct answer: A. Controls placement and organization of output sections for supported toolchains

A linker script describes output section layout, addresses, and placement rules for a supporting toolchain. Embedded systems often need explicit memory layouts.

Correct answer: B. Maps program segments and prepares initial execution state

An executable loader maps program segments into memory, performs required relocations or library setup, and establishes the initial execution state.

Correct answer: C. Expresses target machine instructions as binary fields

Instruction encoding translates a chosen machine operation into the architecture’s prescribed bit fields, such as opcode and operand representations.

Correct answer: D. Locally replaces a machine instruction pattern with an equivalent improvement

A peephole rewrite substitutes a short local instruction sequence with an equivalent and potentially cheaper one. The rewrite must preserve the required flags and effects.

Correct answer: A. Uses target latency registers and instruction constraints

Machine-specific optimization considers the target processor’s registers, instruction latencies, scheduling rules and addressing modes to improve generated code.

Correct answer: B. Maps generated code back to source constructs for debugging

Debug metadata relates generated instruction addresses and machine locations to source lines, names and types. This helps tools inspect optimized programs.

Correct answer: C. Associates transformed locations with original source locations

Source maps connect locations in generated or transformed code to locations in original source files. They are especially useful for transpiled or minified programs.

Correct answer: D. Reports errors warnings or observations with useful source context

Diagnostics report compilation errors, warnings or notes with useful context such as source positions. Precise explanations help programmers identify and repair problems.

Correct answer: A. Rebuilds affected program parts after changes instead of always rebuilding all modules

Incremental compilation rebuilds only program regions affected by edits and dependent changes. Correct dependency tracking prevents using stale compiled artifacts.

Correct answer: B. Tracks relationships that determine what must rebuild

A dependency graph represents which build outputs depend on which inputs or targets. It determines rebuild order and what changes require recompilation.

Correct answer: C. Uses a compiler implementation to help build a compiler for the same or related language

Bootstrapping builds a compiler with an existing compiler, often for its own implementation language. A self-hosting compiler can eventually compile newer versions of itself.

Correct answer: D. Checks that generated code respects intended language semantics on representative cases

Validation tests check that generated code behaves according to intended semantics on representative cases. Testing can detect miscompilations but does not prove universal correctness.