Clean up external users of GlobalValue::getGUID(StringRef) (#129644)

See https://discourse.llvm.org/t/rfc-keep-globalvalue-guids-stable/84801
for context.

This is a non-functional change which just changes the interface of
GlobalValue, in preparation for future functional changes. This part
touches a fair few users, so is split out for ease of review. Future
changes to the GlobalValue implementation can then be focused purely on
that class.

This does the following:

* Rename GlobalValue::getGUID(StringRef) to
  getGUIDAssumingExternalLinkage. This is simply making explicit at the
  callsite what is currently implicit.
* Where possible, migrate users to directly calling getGUID on a
  GlobalValue instance.
* Otherwise, where possible, have them call the newly renamed
  getGUIDAssumingExternalLinkage, to make the assumption explicit.


There are a few cases where neither of the above are possible, as the
caller saves and reconstructs the necessary information to compute the
GUID themselves. We want to migrate these callers eventually, but for
this first step we leave them be.

Author: orodley

bolt/lib/Rewrite/PseudoProbeRewriter.cpp

@@ -147,7 +147,7 @@ void PseudoProbeRewriter::parsePseudoProbe(bool ProfiledOnly) {
         if (!Name)
           continue;
         SymName = *Name;
-        uint64_t GUID = Function::getGUID(SymName);
+        uint64_t GUID = Function::getGUIDAssumingExternalLinkage(SymName);
         FuncStartAddrs[GUID] = F->getAddress();
         if (ProfiledOnly && HasProfile)
           GuidFilter.insert(GUID);
@@ -173,7 +173,7 @@ void PseudoProbeRewriter::parsePseudoProbe(bool ProfiledOnly) {
   const GUIDProbeFunctionMap &GUID2Func = ProbeDecoder.getGUID2FuncDescMap();
   // Checks GUID in GUID2Func and returns it if it's present or null otherwise.
   auto checkGUID = [&](StringRef SymName) -> uint64_t {
-    uint64_t GUID = Function::getGUID(SymName);
+    uint64_t GUID = Function::getGUIDAssumingExternalLinkage(SymName);
     if (GUID2Func.find(GUID) == GUID2Func.end())
       return 0;
     return GUID;
@@ -435,7 +435,7 @@ void PseudoProbeRewriter::encodePseudoProbes() {
     for (const BinaryFunction *F : BC.getAllBinaryFunctions()) {
       const uint64_t Addr =
           F->isEmitted() ? F->getOutputAddress() : F->getAddress();
-      FuncStartAddrs[Function::getGUID(
+      FuncStartAddrs[Function::getGUIDAssumingExternalLinkage(
           NameResolver::restore(F->getOneName()))] = Addr;
     }
     DummyDecoder.buildAddress2ProbeMap(

llvm/examples/OrcV2Examples/LLJITWithThinLTOSummaries/LLJITWithThinLTOSummaries.cpp

@@ -77,7 +77,9 @@ class DuplicateDefinitionInSummary
 
   void log(raw_ostream &OS) const override {
     OS << "Duplicate symbol for global value '" << GlobalValueName
-       << "' (GUID: " << GlobalValue::getGUID(GlobalValueName) << ") in:\n";
+       << "' (GUID: "
+       << GlobalValue::getGUIDAssumingExternalLinkage(GlobalValueName)
+       << ") in:\n";
     for (const std::string &Path : ModulePaths) {
       OS << "    " << Path << "\n";
     }
@@ -110,8 +112,9 @@ class DefinitionNotFoundInSummary
   }
 
   void log(raw_ostream &OS) const override {
-    OS << "No symbol for global value '" << GlobalValueName
-       << "' (GUID: " << GlobalValue::getGUID(GlobalValueName) << ") in:\n";
+    OS << "No symbol for global value '" << GlobalValueName << "' (GUID: "
+       << GlobalValue::getGUIDAssumingExternalLinkage(GlobalValueName)
+       << ") in:\n";
     for (const std::string &Path : ModulePaths) {
       OS << "    " << Path << "\n";
     }
@@ -135,7 +138,8 @@ char DefinitionNotFoundInSummary::ID = 0;
 Expected<StringRef> getMainModulePath(StringRef FunctionName,
                                       ModuleSummaryIndex &Index) {
   // Summaries use unmangled names.
-  GlobalValue::GUID G = GlobalValue::getGUID(FunctionName);
+  GlobalValue::GUID G =
+      GlobalValue::getGUIDAssumingExternalLinkage(FunctionName);
   ValueInfo VI = Index.getValueInfo(G);
 
   // We need a unique definition, otherwise don't try further.

llvm/include/llvm/IR/GlobalValue.h

@@ -570,6 +570,11 @@ class GlobalValue : public Constant {
     return Name;
   }
 
+  /// Declare a type to represent a global unique identifier for a global value.
+  /// This is a 64 bits hash that is used by PGO and ThinLTO to have a compact
+  /// unique way to identify a symbol.
+  using GUID = uint64_t;
+
   /// Return the modified name for a global value suitable to be
   /// used as the key for a global lookup (e.g. profile or ThinLTO).
   /// The value's original name is \c Name and has linkage of type
@@ -578,22 +583,22 @@ class GlobalValue : public Constant {
                                          GlobalValue::LinkageTypes Linkage,
                                          StringRef FileName);
 
+private:
   /// Return the modified name for this global value suitable to be
   /// used as the key for a global lookup (e.g. profile or ThinLTO).
   std::string getGlobalIdentifier() const;
 
-  /// Declare a type to represent a global unique identifier for a global value.
-  /// This is a 64 bits hash that is used by PGO and ThinLTO to have a compact
-  /// unique way to identify a symbol.
-  using GUID = uint64_t;
-
-  /// Return a 64-bit global unique ID constructed from global value name
-  /// (i.e. returned by getGlobalIdentifier()).
-  static GUID getGUID(StringRef GlobalName);
+public:
+  /// Return a 64-bit global unique ID constructed from the name of a global
+  /// symbol. Since this call doesn't supply the linkage or defining filename,
+  /// the GUID computation will assume that the global has external linkage.
+  static GUID getGUIDAssumingExternalLinkage(StringRef GlobalName);
 
   /// Return a 64-bit global unique ID constructed from global value name
   /// (i.e. returned by getGlobalIdentifier()).
-  GUID getGUID() const { return getGUID(getGlobalIdentifier()); }
+  GUID getGUID() const {
+    return getGUIDAssumingExternalLinkage(getGlobalIdentifier());
+  }
 
   /// @name Materialization
   /// Materialization is used to construct functions only as they're needed.

llvm/include/llvm/IR/ModuleSummaryIndex.h

@@ -1321,14 +1321,14 @@ class CfiFunctionIndex {
 
   template <typename... Args> void emplace(Args &&...A) {
     StringRef S(std::forward<Args>(A)...);
-    GlobalValue::GUID GUID =
-        GlobalValue::getGUID(GlobalValue::dropLLVMManglingEscape(S));
+    GlobalValue::GUID GUID = GlobalValue::getGUIDAssumingExternalLinkage(
+        GlobalValue::dropLLVMManglingEscape(S));
     Index[GUID].emplace(S);
   }
 
   size_t count(StringRef S) const {
-    GlobalValue::GUID GUID =
-        GlobalValue::getGUID(GlobalValue::dropLLVMManglingEscape(S));
+    GlobalValue::GUID GUID = GlobalValue::getGUIDAssumingExternalLinkage(
+        GlobalValue::dropLLVMManglingEscape(S));
     auto I = Index.find(GUID);
     if (I == Index.end())
       return 0;
@@ -1731,8 +1731,10 @@ class ModuleSummaryIndex {
   /// Add a global value summary for a value of the given name.
   void addGlobalValueSummary(StringRef ValueName,
                              std::unique_ptr<GlobalValueSummary> Summary) {
-    addGlobalValueSummary(getOrInsertValueInfo(GlobalValue::getGUID(ValueName)),
-                          std::move(Summary));
+    addGlobalValueSummary(
+        getOrInsertValueInfo(
+            GlobalValue::getGUIDAssumingExternalLinkage(ValueName)),
+        std::move(Summary));
   }
 
   /// Add a global value summary for the given ValueInfo.
@@ -1869,19 +1871,22 @@ class ModuleSummaryIndex {
   /// This accessor can mutate the map and therefore should not be used in
   /// the ThinLTO backends.
   TypeIdSummary &getOrInsertTypeIdSummary(StringRef TypeId) {
-    auto TidIter = TypeIdMap.equal_range(GlobalValue::getGUID(TypeId));
+    auto TidIter = TypeIdMap.equal_range(
+        GlobalValue::getGUIDAssumingExternalLinkage(TypeId));
     for (auto &[GUID, TypeIdPair] : make_range(TidIter))
       if (TypeIdPair.first == TypeId)
         return TypeIdPair.second;
-    auto It = TypeIdMap.insert({GlobalValue::getGUID(TypeId),
-                                {TypeIdSaver.save(TypeId), TypeIdSummary()}});
+    auto It =
+        TypeIdMap.insert({GlobalValue::getGUIDAssumingExternalLinkage(TypeId),
+                          {TypeIdSaver.save(TypeId), TypeIdSummary()}});
     return It->second.second;
   }
 
   /// This returns either a pointer to the type id summary (if present in the
   /// summary map) or null (if not present). This may be used when importing.
   const TypeIdSummary *getTypeIdSummary(StringRef TypeId) const {
-    auto TidIter = TypeIdMap.equal_range(GlobalValue::getGUID(TypeId));
+    auto TidIter = TypeIdMap.equal_range(
+        GlobalValue::getGUIDAssumingExternalLinkage(TypeId));
     for (const auto &[GUID, TypeIdPair] : make_range(TidIter))
       if (TypeIdPair.first == TypeId)
         return &TypeIdPair.second;

llvm/include/llvm/IR/ModuleSummaryIndexYAML.h

@@ -314,7 +314,7 @@ template <> struct CustomMappingTraits<TypeIdSummaryMapTy> {
   static void inputOne(IO &io, StringRef Key, TypeIdSummaryMapTy &V) {
     TypeIdSummary TId;
     io.mapRequired(Key.str().c_str(), TId);
-    V.insert({GlobalValue::getGUID(Key), {Key, TId}});
+    V.insert({GlobalValue::getGUIDAssumingExternalLinkage(Key), {Key, TId}});
   }
   static void output(IO &io, TypeIdSummaryMapTy &V) {
     for (auto &TidIter : V)

llvm/include/llvm/ProfileData/SampleProf.h

@@ -1299,7 +1299,7 @@ class FunctionSamples {
 static inline FunctionId getRepInFormat(StringRef Name) {
   if (Name.empty() || !FunctionSamples::UseMD5)
     return FunctionId(Name);
-  return FunctionId(Function::getGUID(Name));
+  return FunctionId(Function::getGUIDAssumingExternalLinkage(Name));
 }
 
 raw_ostream &operator<<(raw_ostream &OS, const FunctionSamples &FS);

llvm/include/llvm/Transforms/Utils/SampleProfileLoaderBaseImpl.h

@@ -109,11 +109,12 @@ class PseudoProbeManager {
   }
 
   const PseudoProbeDescriptor *getDesc(StringRef FProfileName) const {
-    return getDesc(Function::getGUID(FProfileName));
+    return getDesc(Function::getGUIDAssumingExternalLinkage(FProfileName));
   }
 
   const PseudoProbeDescriptor *getDesc(const Function &F) const {
-    return getDesc(Function::getGUID(FunctionSamples::getCanonicalFnName(F)));
+    return getDesc(Function::getGUIDAssumingExternalLinkage(
+        FunctionSamples::getCanonicalFnName(F)));
   }
 
   bool profileIsHashMismatched(const PseudoProbeDescriptor &FuncDesc,

llvm/lib/Analysis/CtxProfAnalysis.cpp

@@ -436,7 +436,7 @@ PreservedAnalyses AssignGUIDPass::run(Module &M, ModuleAnalysisManager &MAM) {
 GlobalValue::GUID AssignGUIDPass::getGUID(const Function &F) {
   if (F.isDeclaration()) {
     assert(GlobalValue::isExternalLinkage(F.getLinkage()));
-    return GlobalValue::getGUID(F.getGlobalIdentifier());
+    return F.getGUID();
   }
   auto *MD = F.getMetadata(GUIDMetadataName);
   assert(MD && "guid not found for defined function");

llvm/lib/Analysis/ModuleSummaryAnalysis.cpp

@@ -222,7 +222,8 @@ static void addIntrinsicToSummary(
     auto *TypeId = dyn_cast<MDString>(TypeMDVal->getMetadata());
     if (!TypeId)
       break;
-    GlobalValue::GUID Guid = GlobalValue::getGUID(TypeId->getString());
+    GlobalValue::GUID Guid =
+        GlobalValue::getGUIDAssumingExternalLinkage(TypeId->getString());
 
     // Produce a summary from type.test intrinsics. We only summarize type.test
     // intrinsics that are used other than by an llvm.assume intrinsic.
@@ -250,7 +251,8 @@ static void addIntrinsicToSummary(
     auto *TypeId = dyn_cast<MDString>(TypeMDVal->getMetadata());
     if (!TypeId)
       break;
-    GlobalValue::GUID Guid = GlobalValue::getGUID(TypeId->getString());
+    GlobalValue::GUID Guid =
+        GlobalValue::getGUIDAssumingExternalLinkage(TypeId->getString());
 
     SmallVector<DevirtCallSite, 4> DevirtCalls;
     SmallVector<Instruction *, 4> LoadedPtrs;
@@ -904,7 +906,8 @@ static void computeAliasSummary(ModuleSummaryIndex &Index, const GlobalAlias &A,
 
 // Set LiveRoot flag on entries matching the given value name.
 static void setLiveRoot(ModuleSummaryIndex &Index, StringRef Name) {
-  if (ValueInfo VI = Index.getValueInfo(GlobalValue::getGUID(Name)))
+  if (ValueInfo VI =
+          Index.getValueInfo(GlobalValue::getGUIDAssumingExternalLinkage(Name)))
     for (const auto &Summary : VI.getSummaryList())
       Summary->setLive(true);
 }

llvm/lib/AsmParser/LLParser.cpp

@@ -9049,7 +9049,7 @@ bool LLParser::parseTypeIdEntry(unsigned ID) {
     for (auto TIDRef : FwdRefTIDs->second) {
       assert(!*TIDRef.first &&
              "Forward referenced type id GUID expected to be 0");
-      *TIDRef.first = GlobalValue::getGUID(Name);
+      *TIDRef.first = GlobalValue::getGUIDAssumingExternalLinkage(Name);
     }
     ForwardRefTypeIds.erase(FwdRefTIDs);
   }
@@ -9154,7 +9154,7 @@ bool LLParser::parseTypeIdCompatibleVtableEntry(unsigned ID) {
     for (auto TIDRef : FwdRefTIDs->second) {
       assert(!*TIDRef.first &&
              "Forward referenced type id GUID expected to be 0");
-      *TIDRef.first = GlobalValue::getGUID(Name);
+      *TIDRef.first = GlobalValue::getGUIDAssumingExternalLinkage(Name);
     }
     ForwardRefTypeIds.erase(FwdRefTIDs);
   }
@@ -9470,7 +9470,7 @@ bool LLParser::addGlobalValueToIndex(
       assert(
           (!GlobalValue::isLocalLinkage(Linkage) || !SourceFileName.empty()) &&
           "Need a source_filename to compute GUID for local");
-      GUID = GlobalValue::getGUID(
+      GUID = GlobalValue::getGUIDAssumingExternalLinkage(
           GlobalValue::getGlobalIdentifier(Name, Linkage, SourceFileName));
       VI = Index->getOrInsertValueInfo(GUID, Index->saveString(Name));
     }