[clang] nonblocking/nonallocating attributes: 2nd pass caller/callee analysis (PR #99656)
via cfe-commits
cfe-commits at lists.llvm.org
Thu Jul 25 20:45:13 PDT 2024
================
@@ -2397,6 +2397,1262 @@ class UnsafeBufferUsageReporter : public UnsafeBufferUsageHandler {
};
} // namespace
+// =============================================================================
+
+namespace FXAnalysis {
+
+enum class DiagnosticID : uint8_t {
+ None = 0, // sentinel for an empty Diagnostic
+ Throws,
+ Catches,
+ CallsObjC,
+ AllocatesMemory,
+ HasStaticLocal,
+ AccessesThreadLocal,
+
+ // These only apply to callees, where the analysis stops at the Decl
+ DeclDisallowsInference,
+
+ CallsDeclWithoutEffect,
+ CallsExprWithoutEffect,
+};
+
+// Holds an effect diagnosis, potentially for the entire duration of the
+// analysis phase, in order to refer to it when explaining why a caller has been
+// made unsafe by a callee.
+struct Diagnostic {
+ FunctionEffect Effect;
+ DiagnosticID ID = DiagnosticID::None;
+ SourceLocation Loc;
+ const Decl *Callee = nullptr; // only valid for Calls*
+
+ Diagnostic() = default;
+
+ Diagnostic(const FunctionEffect &Effect, DiagnosticID ID, SourceLocation Loc,
+ const Decl *Callee = nullptr)
+ : Effect(Effect), ID(ID), Loc(Loc), Callee(Callee) {}
+};
+
+enum class SpecialFuncType : uint8_t { None, OperatorNew, OperatorDelete };
+enum class CallType {
+ // unknown: probably function pointer
+ Unknown,
+ Function,
+ Virtual,
+ Block
+};
+
+// Return whether a function's effects CAN be verified.
+// The question of whether it SHOULD be verified is independent.
+static bool functionIsVerifiable(const FunctionDecl *FD) {
+ if (!(FD->hasBody() || FD->isInlined())) {
+ // externally defined; we couldn't verify if we wanted to.
+ return false;
+ }
+ if (FD->isTrivial()) {
+ // Otherwise `struct x { int a; };` would have an unverifiable default
+ // constructor.
+ return true;
+ }
+ return true;
+}
+
+/// A mutable set of FunctionEffect, for use in places where any conditions
+/// have been resolved or can be ignored.
+class EffectSet {
+ // This implementation optimizes footprint, since we hold one of these for
+ // every function visited, which, due to inference, can be many more functions
+ // than have declared effects.
+
+ template <typename T, typename SizeT, SizeT Capacity> struct FixedVector {
+ SizeT Count = 0;
+ T Items[Capacity] = {};
+
+ using value_type = T;
+
+ using iterator = T *;
+ using const_iterator = const T *;
+ iterator begin() { return &Items[0]; }
+ iterator end() { return &Items[Count]; }
+ const_iterator begin() const { return &Items[0]; }
+ const_iterator end() const { return &Items[Count]; }
+ const_iterator cbegin() const { return &Items[0]; }
+ const_iterator cend() const { return &Items[Count]; }
+
+ void insert(iterator I, const T &Value) {
+ assert(Count < Capacity);
+ iterator E = end();
+ if (I != E)
+ std::copy_backward(I, E, E + 1);
+ *I = Value;
+ ++Count;
+ }
+
+ void push_back(const T &Value) {
+ assert(Count < Capacity);
+ Items[Count++] = Value;
+ }
+ };
+
+ // As long as FunctionEffect is only 1 byte, and there are only 2 verifiable
+ // effects, this fixed-size vector with a capacity of 7 is more than
+ // sufficient and is only 8 bytes.
+ FixedVector<FunctionEffect, uint8_t, 7> Impl;
+
+public:
+ EffectSet() = default;
+ explicit EffectSet(FunctionEffectsRef FX) { insert(FX); }
+
+ operator ArrayRef<FunctionEffect>() const {
+ return ArrayRef(Impl.cbegin(), Impl.cend());
+ }
+
+ using iterator = const FunctionEffect *;
+ iterator begin() const { return Impl.cbegin(); }
+ iterator end() const { return Impl.cend(); }
+
+ void insert(const FunctionEffect &Effect) {
+ FunctionEffect *Iter = Impl.begin();
+ FunctionEffect *End = Impl.end();
+ // linear search; lower_bound is overkill for a tiny vector like this
+ for (; Iter != End; ++Iter) {
+ if (*Iter == Effect)
+ return;
+ if (Effect < *Iter)
+ break;
+ }
+ Impl.insert(Iter, Effect);
+ }
+ void insert(const EffectSet &Set) {
+ for (const FunctionEffect &Item : Set) {
+ // push_back because set is already sorted
+ Impl.push_back(Item);
+ }
+ }
+ void insert(FunctionEffectsRef FX) {
+ for (const FunctionEffectWithCondition &EC : FX) {
+ assert(EC.Cond.getCondition() ==
+ nullptr); // should be resolved by now, right?
+ // push_back because set is already sorted
+ Impl.push_back(EC.Effect);
+ }
+ }
+ bool contains(const FunctionEffect::Kind EK) const {
+ for (const FunctionEffect &E : Impl)
+ if (E.kind() == EK)
+ return true;
+ return false;
+ }
+
+ void dump(llvm::raw_ostream &OS) const;
+
+ static EffectSet difference(ArrayRef<FunctionEffect> LHS,
+ ArrayRef<FunctionEffect> RHS) {
+ EffectSet Result;
+ std::set_difference(LHS.begin(), LHS.end(), RHS.begin(), RHS.end(),
+ std::back_inserter(Result.Impl));
+ return Result;
+ }
+};
+
+LLVM_DUMP_METHOD void EffectSet::dump(llvm::raw_ostream &OS) const {
+ OS << "Effects{";
+ bool First = true;
+ for (const FunctionEffect &Effect : *this) {
+ if (!First)
+ OS << ", ";
+ else
+ First = false;
+ OS << Effect.name();
+ }
+ OS << "}";
+}
+
+// Transitory, more extended information about a callable, which can be a
+// function, block, function pointer, etc.
+struct CallableInfo {
+ // CDecl holds the function's definition, if any.
+ // FunctionDecl if CallType::Function or Virtual
+ // BlockDecl if CallType::Block
+ const Decl *CDecl;
+ mutable std::optional<std::string> MaybeName;
+ SpecialFuncType FuncType = SpecialFuncType::None;
+ EffectSet Effects;
+ CallType CType = CallType::Unknown;
+
+ CallableInfo(Sema &SemaRef, const Decl &CD,
+ SpecialFuncType FT = SpecialFuncType::None)
+ : CDecl(&CD), FuncType(FT) {
+ FunctionEffectsRef FXRef;
+
+ if (auto *FD = dyn_cast<FunctionDecl>(CDecl)) {
+ // Use the function's definition, if any.
+ if (const FunctionDecl *Def = FD->getDefinition())
+ CDecl = FD = Def;
+ CType = CallType::Function;
+ if (auto *Method = dyn_cast<CXXMethodDecl>(FD);
+ Method && Method->isVirtual())
+ CType = CallType::Virtual;
+ FXRef = FD->getFunctionEffects();
+ } else if (auto *BD = dyn_cast<BlockDecl>(CDecl)) {
+ CType = CallType::Block;
+ FXRef = BD->getFunctionEffects();
+ } else if (auto *VD = dyn_cast<ValueDecl>(CDecl)) {
+ // ValueDecl is function, enum, or variable, so just look at its type.
+ FXRef = FunctionEffectsRef::get(VD->getType());
+ }
+ Effects = EffectSet(FXRef);
+ }
+
+ bool isDirectCall() const {
+ return CType == CallType::Function || CType == CallType::Block;
+ }
+
+ bool isVerifiable() const {
+ switch (CType) {
+ case CallType::Unknown:
+ case CallType::Virtual:
+ break;
+ case CallType::Block:
+ return true;
+ case CallType::Function:
+ return functionIsVerifiable(dyn_cast<FunctionDecl>(CDecl));
+ }
+ return false;
+ }
+
+ /// Generate a name for logging and diagnostics.
+ std::string name(Sema &Sem) const {
+ if (!MaybeName) {
+ std::string Name;
+ llvm::raw_string_ostream OS(Name);
+
+ if (auto *FD = dyn_cast<FunctionDecl>(CDecl))
+ FD->getNameForDiagnostic(OS, Sem.getPrintingPolicy(),
+ /*Qualified=*/true);
+ else if (auto *BD = dyn_cast<BlockDecl>(CDecl))
+ OS << "(block " << BD->getBlockManglingNumber() << ")";
+ else if (auto *VD = dyn_cast<NamedDecl>(CDecl))
+ VD->printQualifiedName(OS);
+ MaybeName = Name;
+ }
+ return *MaybeName;
+ }
+};
+
+// ----------
+// Map effects to single diagnostics, to hold the first (of potentially many)
+// diagnostics pertaining to an effect, per function.
+class EffectToDiagnosticMap {
+ // Since we currently only have a tiny number of effects (typically no more
+ // than 1), use a sorted SmallVector with an inline capacity of 1. Since it
+ // is often empty, use a unique_ptr to the SmallVector.
+ // Note that Diagnostic itself contains a FunctionEffect which is the key.
+ using ImplVec = llvm::SmallVector<Diagnostic, 1>;
+ std::unique_ptr<ImplVec> Impl;
+
+public:
+ // Insert a new diagnostic if we do not already have one for its effect.
+ void maybeInsert(const Diagnostic &Diag) {
+ if (Impl == nullptr)
+ Impl = std::make_unique<ImplVec>();
+ auto *Iter = _find(Diag.Effect);
+ if (Iter != Impl->end() && Iter->Effect == Diag.Effect)
+ return;
+
+ Impl->insert(Iter, Diag);
+ }
+
+ const Diagnostic *lookup(FunctionEffect Key) {
+ if (Impl == nullptr)
+ return nullptr;
+
+ auto *Iter = _find(Key);
+ if (Iter != Impl->end() && Iter->Effect == Key)
+ return &*Iter;
+
+ return nullptr;
+ }
+
+ size_t size() const { return Impl ? Impl->size() : 0; }
+
+private:
+ ImplVec::iterator _find(const FunctionEffect &key) {
+ // A linear search suffices for a tiny number of possible effects.
+ auto *End = Impl->end();
+ for (auto *Iter = Impl->begin(); Iter != End; ++Iter)
+ if (!(Iter->Effect < key))
+ return Iter;
+ return End;
+ }
+};
+
+// ----------
+// State pertaining to a function whose AST is walked and whose effect analysis
+// is dependent on a subsequent analysis of other functions.
+class PendingFunctionAnalysis {
+ friend class CompleteFunctionAnalysis;
+
+public:
+ struct DirectCall {
+ const Decl *Callee;
+ SourceLocation CallLoc;
+ // Not all recursive calls are detected, just enough
+ // to break cycles.
+ bool Recursed = false;
+
+ DirectCall(const Decl *D, SourceLocation CallLoc)
+ : Callee(D), CallLoc(CallLoc) {}
+ };
+
+ // We always have two disjoint sets of effects to verify:
+ // 1. Effects declared explicitly by this function.
+ // 2. All other inferrable effects needing verification.
+ EffectSet DeclaredVerifiableEffects;
+ EffectSet FXToInfer;
+
+private:
+ // Diagnostics pertaining to the function's explicit effects.
+ SmallVector<Diagnostic, 0> DiagnosticsForExplicitFX;
+
+ // Diagnostics pertaining to other, non-explicit, inferrable effects.
+ EffectToDiagnosticMap InferrableEffectToFirstDiagnostic;
+
+ // These unverified direct calls are what keeps the analysis "pending",
+ // until the callees can be verified.
+ SmallVector<DirectCall, 0> UnverifiedDirectCalls;
+
+public:
+ PendingFunctionAnalysis(
+ Sema &Sem, const CallableInfo &CInfo,
+ ArrayRef<FunctionEffect> AllInferrableEffectsToVerify) {
+ DeclaredVerifiableEffects = CInfo.Effects;
+
+ // Check for effects we are not allowed to infer
+ EffectSet InferrableFX;
+
+ for (const FunctionEffect &effect : AllInferrableEffectsToVerify) {
+ if (effect.canInferOnFunction(*CInfo.CDecl))
+ InferrableFX.insert(effect);
+ else {
+ // Add a diagnostic for this effect if a caller were to
+ // try to infer it.
+ InferrableEffectToFirstDiagnostic.maybeInsert(
+ Diagnostic(effect, DiagnosticID::DeclDisallowsInference,
+ CInfo.CDecl->getLocation()));
+ }
+ }
+ // InferrableFX is now the set of inferrable effects which are not
+ // prohibited
+ FXToInfer = EffectSet::difference(InferrableFX, DeclaredVerifiableEffects);
+ }
+
+ // Hide the way that diagnostics for explicitly required effects vs. inferred
+ // ones are handled differently.
+ void checkAddDiagnostic(bool Inferring, const Diagnostic &NewDiag) {
+ if (!Inferring)
+ DiagnosticsForExplicitFX.push_back(NewDiag);
+ else
+ InferrableEffectToFirstDiagnostic.maybeInsert(NewDiag);
+ }
+
+ void addUnverifiedDirectCall(const Decl *D, SourceLocation CallLoc) {
+ UnverifiedDirectCalls.emplace_back(D, CallLoc);
+ }
+
+ // Analysis is complete when there are no unverified direct calls.
+ bool isComplete() const { return UnverifiedDirectCalls.empty(); }
+
+ const Diagnostic *diagnosticForInferrableEffect(FunctionEffect effect) {
+ return InferrableEffectToFirstDiagnostic.lookup(effect);
+ }
+
+ SmallVector<DirectCall, 0> &unverifiedCalls() {
+ assert(!isComplete());
+ return UnverifiedDirectCalls;
+ }
+
+ SmallVector<Diagnostic, 0> &getDiagnosticsForExplicitFX() {
+ return DiagnosticsForExplicitFX;
+ }
+
+ void dump(Sema &SemaRef, llvm::raw_ostream &OS) const {
+ OS << "Pending: Declared ";
+ DeclaredVerifiableEffects.dump(OS);
+ OS << ", " << DiagnosticsForExplicitFX.size() << " diags; ";
+ OS << " Infer ";
+ FXToInfer.dump(OS);
+ OS << ", " << InferrableEffectToFirstDiagnostic.size() << " diags";
+ if (!UnverifiedDirectCalls.empty()) {
+ OS << "; Calls: ";
+ for (const DirectCall &Call : UnverifiedDirectCalls) {
+ CallableInfo CI(SemaRef, *Call.Callee);
+ OS << " " << CI.name(SemaRef);
+ }
+ }
+ OS << "\n";
+ }
+};
+
+// ----------
+class CompleteFunctionAnalysis {
+ // Current size: 2 pointers
+public:
+ // Has effects which are both the declared ones -- not to be inferred -- plus
+ // ones which have been successfully inferred. These are all considered
+ // "verified" for the purposes of callers; any issue with verifying declared
+ // effects has already been reported and is not the problem of any caller.
+ EffectSet VerifiedEffects;
+
+private:
+ // This is used to generate notes about failed inference.
+ EffectToDiagnosticMap InferrableEffectToFirstDiagnostic;
+
+public:
+ // The incoming Pending analysis is consumed (member(s) are moved-from).
+ CompleteFunctionAnalysis(
+ ASTContext &Ctx, PendingFunctionAnalysis &Pending,
+ const EffectSet &DeclaredEffects,
+ ArrayRef<FunctionEffect> AllInferrableEffectsToVerify) {
+ VerifiedEffects.insert(DeclaredEffects);
+ for (const FunctionEffect &effect : AllInferrableEffectsToVerify)
+ if (Pending.diagnosticForInferrableEffect(effect) == nullptr)
+ VerifiedEffects.insert(effect);
+
+ InferrableEffectToFirstDiagnostic =
+ std::move(Pending.InferrableEffectToFirstDiagnostic);
+ }
+
+ const Diagnostic *firstDiagnosticForEffect(const FunctionEffect &Effect) {
+ return InferrableEffectToFirstDiagnostic.lookup(Effect);
+ }
+
+ void dump(llvm::raw_ostream &OS) const {
+ OS << "Complete: Verified ";
+ VerifiedEffects.dump(OS);
+ OS << "; Infer ";
+ OS << InferrableEffectToFirstDiagnostic.size() << " diags\n";
+ }
+};
+
+const Decl *CanonicalFunctionDecl(const Decl *D) {
+ if (auto *FD = dyn_cast<FunctionDecl>(D)) {
+ FD = FD->getCanonicalDecl();
+ assert(FD != nullptr);
+ return FD;
+ }
+ return D;
+}
+
+// ==========
+class Analyzer {
+ constexpr static int DebugLogLevel = 0;
+ // --
+ Sema &Sem;
+
+ // Subset of Sema.AllEffectsToVerify
+ EffectSet AllInferrableEffectsToVerify;
+
+ using FuncAnalysisPtr =
+ llvm::PointerUnion<PendingFunctionAnalysis *, CompleteFunctionAnalysis *>;
+
+ // Map all Decls analyzed to FuncAnalysisPtr. Pending state is larger
+ // than complete state, so use different objects to represent them.
+ // The state pointers are owned by the container.
+ class AnalysisMap : protected llvm::DenseMap<const Decl *, FuncAnalysisPtr> {
+ using Base = llvm::DenseMap<const Decl *, FuncAnalysisPtr>;
+
+ public:
+ ~AnalysisMap();
+
+ // Use non-public inheritance in order to maintain the invariant
+ // that lookups and insertions are via the canonical Decls.
+
+ FuncAnalysisPtr lookup(const Decl *Key) const {
+ return Base::lookup(CanonicalFunctionDecl(Key));
+ }
+
+ FuncAnalysisPtr &operator[](const Decl *Key) {
+ return Base::operator[](CanonicalFunctionDecl(Key));
+ }
+
+ /// Shortcut for the case where we only care about completed analysis.
+ CompleteFunctionAnalysis *completedAnalysisForDecl(const Decl *D) const {
+ if (FuncAnalysisPtr AP = lookup(D);
+ isa_and_nonnull<CompleteFunctionAnalysis *>(AP))
+ return AP.get<CompleteFunctionAnalysis *>();
+ return nullptr;
+ }
+
+ void dump(Sema &SemaRef, llvm::raw_ostream &OS) {
+ OS << "\nAnalysisMap:\n";
+ for (const auto &item : *this) {
+ CallableInfo CI(SemaRef, *item.first);
+ const auto AP = item.second;
+ OS << item.first << " " << CI.name(SemaRef) << " : ";
+ if (AP.isNull())
+ OS << "null\n";
+ else if (isa<CompleteFunctionAnalysis *>(AP)) {
+ auto *CFA = AP.get<CompleteFunctionAnalysis *>();
+ OS << CFA << " ";
+ CFA->dump(OS);
+ } else if (isa<PendingFunctionAnalysis *>(AP)) {
+ auto *PFA = AP.get<PendingFunctionAnalysis *>();
+ OS << PFA << " ";
+ PFA->dump(SemaRef, OS);
+ } else
+ llvm_unreachable("never");
+ }
+ OS << "---\n";
+ }
+ };
+ AnalysisMap DeclAnalysis;
+
+public:
+ Analyzer(Sema &S) : Sem(S) {}
+
+ void run(const TranslationUnitDecl &TU) {
+ // Gather all of the effects to be verified to see what operations need to
+ // be checked, and to see which ones are inferrable.
+ for (const FunctionEffectWithCondition &CFE : Sem.AllEffectsToVerify) {
+ const FunctionEffect &Effect = CFE.Effect;
+ const FunctionEffect::Flags Flags = Effect.flags();
+ if (Flags & FunctionEffect::FE_InferrableOnCallees)
+ AllInferrableEffectsToVerify.insert(Effect);
+ }
+ if constexpr (DebugLogLevel > 0) {
+ llvm::outs() << "AllInferrableEffectsToVerify: ";
+ AllInferrableEffectsToVerify.dump(llvm::outs());
+ llvm::outs() << "\n";
+ }
+
+ // We can use DeclsWithEffectsToVerify as a stack for a
+ // depth-first traversal; there's no need for a second container. But first,
+ // reverse it, so when working from the end, Decls are verified in the order
+ // they are declared.
+ SmallVector<const Decl *> &VerificationQueue = Sem.DeclsWithEffectsToVerify;
+ std::reverse(VerificationQueue.begin(), VerificationQueue.end());
+
+ while (!VerificationQueue.empty()) {
+ const Decl *D = VerificationQueue.back();
+ if (FuncAnalysisPtr AP = DeclAnalysis.lookup(D)) {
+ if (isa<CompleteFunctionAnalysis *>(AP)) {
+ // already done
+ VerificationQueue.pop_back();
+ continue;
+ }
+ if (isa<PendingFunctionAnalysis *>(AP)) {
+ // All children have been traversed; finish analysis.
+ auto *Pending = AP.get<PendingFunctionAnalysis *>();
+ finishPendingAnalysis(D, Pending);
+ VerificationQueue.pop_back();
+ continue;
+ }
+ llvm_unreachable("unexpected DeclAnalysis item");
+ }
+
+ // Not previously visited; begin a new analysis for this Decl.
+ PendingFunctionAnalysis *Pending = verifyDecl(D);
+ if (Pending == nullptr) {
+ // completed now
+ VerificationQueue.pop_back();
+ continue;
+ }
+
+ // Analysis remains pending because there are direct callees to be
+ // verified first. Push them onto the queue.
+ for (PendingFunctionAnalysis::DirectCall &Call :
+ Pending->unverifiedCalls()) {
+ FuncAnalysisPtr AP = DeclAnalysis.lookup(Call.Callee);
+ if (AP.isNull()) {
+ VerificationQueue.push_back(Call.Callee);
+ continue;
+ }
+ if (isa<PendingFunctionAnalysis *>(AP)) {
+ // This indicates recursion (not necessarily direct). For the
+ // purposes of effect analysis, we can just ignore it since
+ // no effects forbid recursion.
+ Call.Recursed = true;
+ continue;
+ }
+ llvm_unreachable("unexpected DeclAnalysis item");
+ }
+ }
+ }
+
+private:
+ // Verify a single Decl. Return the pending structure if that was the result,
+ // else null. This method must not recurse.
+ PendingFunctionAnalysis *verifyDecl(const Decl *D) {
+ CallableInfo CInfo(Sem, *D);
+ bool isExternC = false;
+
+ if (const FunctionDecl *FD = dyn_cast<FunctionDecl>(D)) {
+ assert(FD->getBuiltinID() == 0);
+ isExternC = FD->getCanonicalDecl()->isExternCContext();
+ }
+
+ // For C++, with non-extern "C" linkage only - if any of the Decl's declared
+ // effects forbid throwing (e.g. nonblocking) then the function should also
+ // be declared noexcept.
+ if (Sem.getLangOpts().CPlusPlus && !isExternC) {
+ for (const FunctionEffect &Effect : CInfo.Effects) {
+ if (!(Effect.flags() & FunctionEffect::FE_ExcludeThrow))
+ continue;
+
+ bool IsNoexcept = false;
+ if (auto *FD = D->getAsFunction()) {
+ IsNoexcept = isNoexcept(FD);
+ } else if (auto *BD = dyn_cast<BlockDecl>(D)) {
+ if (auto *TSI = BD->getSignatureAsWritten()) {
+ auto *FPT = TSI->getType()->getAs<FunctionProtoType>();
+ IsNoexcept = FPT->isNothrow() || BD->hasAttr<NoThrowAttr>();
+ }
+ }
+ if (!IsNoexcept)
+ Sem.Diag(D->getBeginLoc(),
+ diag::warn_perf_constraint_implies_noexcept)
+ << Effect.name();
+ break;
+ }
+ }
+
+ // Build a PendingFunctionAnalysis on the stack. If it turns out to be
+ // complete, we'll have avoided a heap allocation; if it's incomplete, it's
+ // a fairly trivial move to a heap-allocated object.
+ PendingFunctionAnalysis FAnalysis(Sem, CInfo, AllInferrableEffectsToVerify);
+
+ if constexpr (DebugLogLevel > 0) {
+ llvm::outs() << "\nVerifying " << CInfo.name(Sem) << " ";
+ FAnalysis.dump(Sem, llvm::outs());
+ }
+
+ FunctionBodyASTVisitor Visitor(*this, FAnalysis, CInfo);
+
+ Visitor.run();
+ if (FAnalysis.isComplete()) {
+ completeAnalysis(CInfo, FAnalysis);
+ return nullptr;
+ }
+ // Move the pending analysis to the heap and save it in the map.
+ PendingFunctionAnalysis *PendingPtr =
+ new PendingFunctionAnalysis(std::move(FAnalysis));
+ DeclAnalysis[D] = PendingPtr;
+ if constexpr (DebugLogLevel > 0) {
+ llvm::outs() << "inserted pending " << PendingPtr << "\n";
+ DeclAnalysis.dump(Sem, llvm::outs());
+ }
+ return PendingPtr;
+ }
+
+ // Consume PendingFunctionAnalysis, create with it a CompleteFunctionAnalysis,
+ // inserted in the container.
+ void completeAnalysis(const CallableInfo &CInfo,
+ PendingFunctionAnalysis &Pending) {
+ if (SmallVector<Diagnostic, 0> &Diags =
+ Pending.getDiagnosticsForExplicitFX();
+ !Diags.empty())
+ emitDiagnostics(Diags, CInfo, Sem);
+
+ CompleteFunctionAnalysis *CompletePtr = new CompleteFunctionAnalysis(
+ Sem.getASTContext(), Pending, CInfo.Effects,
+ AllInferrableEffectsToVerify);
+ DeclAnalysis[CInfo.CDecl] = CompletePtr;
+ if constexpr (DebugLogLevel > 0) {
+ llvm::outs() << "inserted complete " << CompletePtr << "\n";
+ DeclAnalysis.dump(Sem, llvm::outs());
+ }
+ }
+
+ // Called after all direct calls requiring inference have been found -- or
+ // not. Repeats calls to FunctionBodyASTVisitor::followCall() but without
+ // the possibility of inference. Deletes Pending.
+ void finishPendingAnalysis(const Decl *D, PendingFunctionAnalysis *Pending) {
+ CallableInfo Caller(Sem, *D);
+ if constexpr (DebugLogLevel > 0) {
+ llvm::outs() << "finishPendingAnalysis for " << Caller.name(Sem) << " : ";
+ Pending->dump(Sem, llvm::outs());
+ llvm::outs() << "\n";
+ }
+ for (const PendingFunctionAnalysis::DirectCall &Call :
+ Pending->unverifiedCalls()) {
+ if (Call.Recursed)
+ continue;
+
+ CallableInfo Callee(Sem, *Call.Callee);
+ followCall(Caller, *Pending, Callee, Call.CallLoc,
+ /*AssertNoFurtherInference=*/true);
+ }
+ completeAnalysis(Caller, *Pending);
+ delete Pending;
+ }
+
+ // Here we have a call to a Decl, either explicitly via a CallExpr or some
+ // other AST construct. PFA pertains to the caller.
+ void followCall(const CallableInfo &Caller, PendingFunctionAnalysis &PFA,
+ const CallableInfo &Callee, SourceLocation CallLoc,
+ bool AssertNoFurtherInference) {
+ const bool DirectCall = Callee.isDirectCall();
+
+ // Initially, the declared effects; inferred effects will be added.
+ EffectSet CalleeEffects = Callee.Effects;
+
+ bool IsInferencePossible = DirectCall;
+
+ if (DirectCall) {
+ if (CompleteFunctionAnalysis *CFA =
+ DeclAnalysis.completedAnalysisForDecl(Callee.CDecl)) {
+ // Combine declared effects with those which may have been inferred.
+ CalleeEffects.insert(CFA->VerifiedEffects);
+ IsInferencePossible = false; // we've already traversed it
+ }
+ }
+
+ if (AssertNoFurtherInference) {
+ assert(!IsInferencePossible);
+ }
+
+ if (!Callee.isVerifiable())
+ IsInferencePossible = false;
+
+ if constexpr (DebugLogLevel > 0) {
+ llvm::outs() << "followCall from " << Caller.name(Sem) << " to "
+ << Callee.name(Sem)
+ << "; verifiable: " << Callee.isVerifiable() << "; callee ";
+ CalleeEffects.dump(llvm::outs());
+ llvm::outs() << "\n";
+ llvm::outs() << " callee " << Callee.CDecl << " canonical "
+ << CanonicalFunctionDecl(Callee.CDecl) << " redecls";
+ for (Decl *D : Callee.CDecl->redecls())
+ llvm::outs() << " " << D;
+
+ llvm::outs() << "\n";
+ }
+
+ auto check1Effect = [&](const FunctionEffect &Effect, bool Inferring) {
+ FunctionEffect::Flags Flags = Effect.flags();
+ bool Diagnose =
+ Effect.shouldDiagnoseFunctionCall(DirectCall, CalleeEffects);
+ if (Diagnose) {
+ // If inference is not allowed, or the target is indirect (virtual
+ // method/function ptr?), generate a diagnostic now.
+ if (!IsInferencePossible ||
+ !(Flags & FunctionEffect::FE_InferrableOnCallees)) {
+ if (Callee.FuncType == SpecialFuncType::None)
+ PFA.checkAddDiagnostic(
+ Inferring, {Effect, DiagnosticID::CallsDeclWithoutEffect,
+ CallLoc, Callee.CDecl});
+ else
+ PFA.checkAddDiagnostic(
+ Inferring, {Effect, DiagnosticID::AllocatesMemory, CallLoc});
+ } else {
+ // Inference is allowed and necessary; defer it.
+ PFA.addUnverifiedDirectCall(Callee.CDecl, CallLoc);
+ }
+ }
+ };
+
+ for (const FunctionEffect &Effect : PFA.DeclaredVerifiableEffects)
+ check1Effect(Effect, false);
+
+ for (const FunctionEffect &Effect : PFA.FXToInfer)
+ check1Effect(Effect, true);
+ }
+
+ // Should only be called when determined to be complete.
+ void emitDiagnostics(SmallVector<Diagnostic, 0> &Diags,
+ const CallableInfo &CInfo, Sema &S) {
+ if (Diags.empty())
+ return;
+ const SourceManager &SM = S.getSourceManager();
+ std::sort(Diags.begin(), Diags.end(),
+ [&SM](const Diagnostic &LHS, const Diagnostic &RHS) {
+ return SM.isBeforeInTranslationUnit(LHS.Loc, RHS.Loc);
+ });
+
+ auto checkAddTemplateNote = [&](const Decl *D) {
+ if (const FunctionDecl *FD = dyn_cast<FunctionDecl>(D)) {
+ while (FD != nullptr && FD->isTemplateInstantiation()) {
+ S.Diag(FD->getPointOfInstantiation(),
+ diag::note_func_effect_from_template);
+ FD = FD->getTemplateInstantiationPattern();
+ }
+ }
+ };
+
+ // Top-level diagnostics are warnings.
+ for (const Diagnostic &Diag : Diags) {
+ StringRef effectName = Diag.Effect.name();
+ switch (Diag.ID) {
+ case DiagnosticID::None:
+ case DiagnosticID::DeclDisallowsInference: // shouldn't happen
+ // here
+ llvm_unreachable("Unexpected diagnostic kind");
+ break;
+ case DiagnosticID::AllocatesMemory:
+ S.Diag(Diag.Loc, diag::warn_func_effect_allocates) << effectName;
+ checkAddTemplateNote(CInfo.CDecl);
+ break;
+ case DiagnosticID::Throws:
+ case DiagnosticID::Catches:
+ S.Diag(Diag.Loc, diag::warn_func_effect_throws_or_catches)
+ << effectName;
+ checkAddTemplateNote(CInfo.CDecl);
+ break;
+ case DiagnosticID::HasStaticLocal:
+ S.Diag(Diag.Loc, diag::warn_func_effect_has_static_local) << effectName;
+ checkAddTemplateNote(CInfo.CDecl);
+ break;
+ case DiagnosticID::AccessesThreadLocal:
+ S.Diag(Diag.Loc, diag::warn_func_effect_uses_thread_local)
+ << effectName;
+ checkAddTemplateNote(CInfo.CDecl);
+ break;
+ case DiagnosticID::CallsObjC:
+ S.Diag(Diag.Loc, diag::warn_func_effect_calls_objc) << effectName;
+ checkAddTemplateNote(CInfo.CDecl);
+ break;
+ case DiagnosticID::CallsExprWithoutEffect:
+ S.Diag(Diag.Loc, diag::warn_func_effect_calls_expr_without_effect)
+ << effectName;
+ checkAddTemplateNote(CInfo.CDecl);
+ break;
+
+ case DiagnosticID::CallsDeclWithoutEffect: {
+ CallableInfo CalleeInfo(S, *Diag.Callee);
+ std::string CalleeName = CalleeInfo.name(S);
+
+ S.Diag(Diag.Loc, diag::warn_func_effect_calls_func_without_effect)
+ << effectName << CalleeName;
+ checkAddTemplateNote(CInfo.CDecl);
+
+ // Emit notes explaining the transitive chain of inferences: Why isn't
+ // the callee safe?
+ for (const Decl *Callee = Diag.Callee; Callee != nullptr;) {
+ std::optional<CallableInfo> MaybeNextCallee;
+ CompleteFunctionAnalysis *Completed =
+ DeclAnalysis.completedAnalysisForDecl(CalleeInfo.CDecl);
+ if (Completed == nullptr) {
+ // No result - could be
+ // - non-inline
+ // - indirect (virtual or through function pointer)
+ // - effect has been explicitly disclaimed (e.g. "blocking")
+ if (CalleeInfo.CType == CallType::Virtual)
+ S.Diag(Callee->getLocation(), diag::note_func_effect_call_virtual)
+ << effectName;
+ else if (CalleeInfo.CType == CallType::Unknown)
+ S.Diag(Callee->getLocation(),
+ diag::note_func_effect_call_func_ptr)
+ << effectName;
+ else if (CalleeInfo.Effects.contains(Diag.Effect.oppositeKind()))
+ S.Diag(Callee->getLocation(),
+ diag::note_func_effect_call_disallows_inference)
+ << effectName;
+ else
+ S.Diag(Callee->getLocation(), diag::note_func_effect_call_extern)
+ << effectName;
+
+ break;
+ }
+ const Diagnostic *PtrDiag2 =
+ Completed->firstDiagnosticForEffect(Diag.Effect);
+ if (PtrDiag2 == nullptr)
+ break;
+
+ const Diagnostic &Diag2 = *PtrDiag2;
+ switch (Diag2.ID) {
+ case DiagnosticID::None:
+ llvm_unreachable("Unexpected diagnostic kind");
+ break;
+ case DiagnosticID::DeclDisallowsInference:
+ S.Diag(Diag2.Loc, diag::note_func_effect_call_disallows_inference)
+ << effectName;
+ break;
+ case DiagnosticID::CallsExprWithoutEffect:
+ S.Diag(Diag2.Loc, diag::note_func_effect_call_func_ptr)
+ << effectName;
+ break;
+ case DiagnosticID::AllocatesMemory:
+ S.Diag(Diag2.Loc, diag::note_func_effect_allocates) << effectName;
+ break;
+ case DiagnosticID::Throws:
+ case DiagnosticID::Catches:
+ S.Diag(Diag2.Loc, diag::note_func_effect_throws_or_catches)
+ << effectName;
+ break;
+ case DiagnosticID::HasStaticLocal:
+ S.Diag(Diag2.Loc, diag::note_func_effect_has_static_local)
+ << effectName;
+ break;
+ case DiagnosticID::AccessesThreadLocal:
+ S.Diag(Diag2.Loc, diag::note_func_effect_uses_thread_local)
+ << effectName;
+ break;
+ case DiagnosticID::CallsObjC:
+ S.Diag(Diag2.Loc, diag::note_func_effect_calls_objc) << effectName;
+ break;
+ case DiagnosticID::CallsDeclWithoutEffect:
+ MaybeNextCallee.emplace(S, *Diag2.Callee);
+ S.Diag(Diag2.Loc, diag::note_func_effect_calls_func_without_effect)
+ << effectName << MaybeNextCallee->name(S);
+ break;
+ }
+ checkAddTemplateNote(Callee);
+ Callee = Diag2.Callee;
+ if (MaybeNextCallee) {
+ CalleeInfo = *MaybeNextCallee;
+ CalleeName = CalleeInfo.name(S);
+ }
+ }
+ } break;
+ }
+ }
+ }
+
+ // ----------
+ // This AST visitor is used to traverse the body of a function during effect
+ // verification. This happens in 2 situations:
+ // [1] The function has declared effects which need to be validated.
+ // [2] The function has not explicitly declared an effect in question, and is
+ // being checked for implicit conformance.
+ //
+ // Diagnostics are always routed to a PendingFunctionAnalysis, which holds
+ // all diagnostic output.
+ //
+ // Q: Currently we create a new RecursiveASTVisitor for every function
+ // analysis. Is it so lightweight that this is OK? It would appear so.
+ struct FunctionBodyASTVisitor
+ : public RecursiveASTVisitor<FunctionBodyASTVisitor> {
+ // The meanings of the boolean values returned by the Visit methods can be
+ // difficult to remember.
+ constexpr static bool Stop = false;
+ constexpr static bool Proceed = true;
----------------
Sirraide wrote:
I don’t think we do this anywhere else (and I have seen pretty much every AST visitor we have because I’ve been refactoring all of them...). You can think of it as RAV functions returning ‘whether traversal should continue’, so I think just `true`/`false` are fine in this case.
https://github.com/llvm/llvm-project/pull/99656
More information about the cfe-commits
mailing list