Improve handling of instantiated thread_local variables in Itanium C++ ABI.

 * Do not initialize these variables when initializing the rest of the
   thread_locals in the TU; they have unordered initialization so they can be
   initialized by themselves.

   This fixes a rejects-valid bug: we would make the per-variable initializer
   function internal, but put it in a comdat keyed off the variable, resulting
   in link errors when the comdat is selected from a different TU (as the per
   TU TLS init function tries to call an init function that does not exist).

 * On Darwin, when we decide that we're not going to emit a thread wrapper
   function at all, demote its linkage to External. Fixes a verifier failure
   on explicit instantiation of a thread_local variable on Darwin.

llvm-svn: 291865

2 files changed, 31 insertions, 14 deletions

diff --git a/clang/lib/CodeGen/CGDeclCXX.cpp b/clang/lib/CodeGen/CGDeclCXX.cpp
index 8d9d0b21bfe..f56e1821693 100644
--- a/clang/lib/CodeGen/CGDeclCXX.cpp
+++ b/clang/lib/CodeGen/CGDeclCXX.cpp
@@ -353,9 +353,6 @@ CodeGenModule::EmitCXXGlobalVarDeclInitFunc(const VarDecl *D,
 
   if (D->getTLSKind()) {
     // FIXME: Should we support init_priority for thread_local?
-    // FIXME: Ideally, initialization of instantiated thread_local static data
-    // members of class templates should not trigger initialization of other
-    // entities in the TU.
     // FIXME: We only need to register one __cxa_thread_atexit function for the
     // entire TU.
     CXXThreadLocalInits.push_back(Fn);
diff --git a/clang/lib/CodeGen/ItaniumCXXABI.cpp b/clang/lib/CodeGen/ItaniumCXXABI.cpp
index b5d90ea59a4..f7a8dd66c52 100644
--- a/clang/lib/CodeGen/ItaniumCXXABI.cpp
+++ b/clang/lib/CodeGen/ItaniumCXXABI.cpp
@@ -2272,7 +2272,21 @@ void ItaniumCXXABI::EmitThreadLocalInitFuncs(
     ArrayRef<llvm::Function *> CXXThreadLocalInits,
     ArrayRef<const VarDecl *> CXXThreadLocalInitVars) {
   llvm::Function *InitFunc = nullptr;
-  if (!CXXThreadLocalInits.empty()) {
+
+  // Separate initializers into those with ordered (or partially-ordered)
+  // initialization and those with unordered initialization.
+  llvm::SmallVector<llvm::Function *, 8> OrderedInits;
+  llvm::SmallDenseMap<const VarDecl *, llvm::Function *> UnorderedInits;
+  for (unsigned I = 0; I != CXXThreadLocalInits.size(); ++I) {
+    if (isTemplateInstantiation(
+            CXXThreadLocalInitVars[I]->getTemplateSpecializationKind()))
+      UnorderedInits[CXXThreadLocalInitVars[I]->getCanonicalDecl()] =
+          CXXThreadLocalInits[I];
+    else
+      OrderedInits.push_back(CXXThreadLocalInits[I]);
+  }
+
+  if (!OrderedInits.empty()) {
     // Generate a guarded initialization function.
     llvm::FunctionType *FTy =
         llvm::FunctionType::get(CGM.VoidTy, /*isVarArg=*/false);
@@ -2289,24 +2303,28 @@ void ItaniumCXXABI::EmitThreadLocalInitFuncs(
     CharUnits GuardAlign = CharUnits::One();
     Guard->setAlignment(GuardAlign.getQuantity());
 
-    CodeGenFunction(CGM)
-        .GenerateCXXGlobalInitFunc(InitFunc, CXXThreadLocalInits,
-                                   Address(Guard, GuardAlign));
+    CodeGenFunction(CGM).GenerateCXXGlobalInitFunc(InitFunc, OrderedInits,
+                                                   Address(Guard, GuardAlign));
     // On Darwin platforms, use CXX_FAST_TLS calling convention.
     if (CGM.getTarget().getTriple().isOSDarwin()) {
       InitFunc->setCallingConv(llvm::CallingConv::CXX_FAST_TLS);
       InitFunc->addFnAttr(llvm::Attribute::NoUnwind);
     }
   }
+
+  // Emit thread wrappers.
   for (const VarDecl *VD : CXXThreadLocals) {
     llvm::GlobalVariable *Var =
         cast<llvm::GlobalVariable>(CGM.GetGlobalValue(CGM.getMangledName(VD)));
+    llvm::Function *Wrapper = getOrCreateThreadLocalWrapper(VD, Var);
 
     // Some targets require that all access to thread local variables go through
     // the thread wrapper.  This means that we cannot attempt to create a thread
     // wrapper or a thread helper.
-    if (isThreadWrapperReplaceable(VD, CGM) && !VD->hasDefinition())
+    if (isThreadWrapperReplaceable(VD, CGM) && !VD->hasDefinition()) {
+      Wrapper->setLinkage(llvm::Function::ExternalLinkage);
       continue;
+    }
 
     // Mangle the name for the thread_local initialization function.
     SmallString<256> InitFnName;
@@ -2322,18 +2340,21 @@ void ItaniumCXXABI::EmitThreadLocalInitFuncs(
     bool InitIsInitFunc = false;
     if (VD->hasDefinition()) {
       InitIsInitFunc = true;
-      if (InitFunc)
+      llvm::Function *InitFuncToUse = InitFunc;
+      if (isTemplateInstantiation(VD->getTemplateSpecializationKind()))
+        InitFuncToUse = UnorderedInits.lookup(VD->getCanonicalDecl());
+      if (InitFuncToUse)
         Init = llvm::GlobalAlias::create(Var->getLinkage(), InitFnName.str(),
-                                         InitFunc);
+                                         InitFuncToUse);
     } else {
       // Emit a weak global function referring to the initialization function.
       // This function will not exist if the TU defining the thread_local
       // variable in question does not need any dynamic initialization for
       // its thread_local variables.
       llvm::FunctionType *FnTy = llvm::FunctionType::get(CGM.VoidTy, false);
-      Init = llvm::Function::Create(
-          FnTy, llvm::GlobalVariable::ExternalWeakLinkage, InitFnName.str(),
-          &CGM.getModule());
+      Init = llvm::Function::Create(FnTy,
+                                    llvm::GlobalVariable::ExternalWeakLinkage,
+                                    InitFnName.str(), &CGM.getModule());
       const CGFunctionInfo &FI = CGM.getTypes().arrangeNullaryFunction();
       CGM.SetLLVMFunctionAttributes(nullptr, FI, cast<llvm::Function>(Init));
     }
@@ -2341,7 +2362,6 @@ void ItaniumCXXABI::EmitThreadLocalInitFuncs(
     if (Init)
       Init->setVisibility(Var->getVisibility());
 
-    llvm::Function *Wrapper = getOrCreateThreadLocalWrapper(VD, Var);
     llvm::LLVMContext &Context = CGM.getModule().getContext();
     llvm::BasicBlock *Entry = llvm::BasicBlock::Create(Context, "", Wrapper);
     CGBuilderTy Builder(CGM, Entry);