[flang] add option to generate runtime type info as external (#146071)

Reland #145901 with a fix for shared library builds.

So far flang generates runtime derived type info global definitions (as
opposed to declarations) for all the types used in the current
compilation unit even when the derived types are defined in other
compilation units. It is using linkonce_odr to achieve derived type
descriptor address "uniqueness" aspect needed to match two derived type
inside the runtime.

This comes at a big compile time cost because of all the extra globals
and their definitions in apps with many and complex derived types.

This patch adds and experimental option to only generate the rtti
definition for the types defined in the current compilation unit and to
only generate external declaration for the derived type descriptor
object of types defined elsewhere.

Note that objects compiled with this option are not compatible with
object files compiled without because files compiled without it may drop
the rtti for type they defined if it is not used in the compilation unit
because of the linkonce_odr aspect.

I am adding the option so that we can better measure the extra cost of
the current approach on apps and allow speeding up some compilation
where devirtualization does not matter (and the build config links to
all module file object anyway).

18 files changed, 267 insertions, 160 deletions

diff --git a/flang/include/flang/Evaluate/tools.h b/flang/include/flang/Evaluate/tools.h
index d26a477..cad1b63 100644
--- a/flang/include/flang/Evaluate/tools.h
+++ b/flang/include/flang/Evaluate/tools.h
@@ -1585,6 +1585,7 @@ bool IsExtensibleType(const DerivedTypeSpec *);
 bool IsSequenceOrBindCType(const DerivedTypeSpec *);
 bool IsBuiltinDerivedType(const DerivedTypeSpec *derived, const char *name);
 bool IsBuiltinCPtr(const Symbol &);
+bool IsFromBuiltinModule(const Symbol &);
 bool IsEventType(const DerivedTypeSpec *);
 bool IsLockType(const DerivedTypeSpec *);
 bool IsNotifyType(const DerivedTypeSpec *);
diff --git a/flang/include/flang/Lower/LoweringOptions.def b/flang/include/flang/Lower/LoweringOptions.def
index d97abf4..3263ab1 100644
--- a/flang/include/flang/Lower/LoweringOptions.def
+++ b/flang/include/flang/Lower/LoweringOptions.def
@@ -66,5 +66,9 @@ ENUM_LOWERINGOPT(RepackArraysWhole, unsigned, 1, 0)
 /// If true, CUDA Fortran runtime check is inserted.
 ENUM_LOWERINGOPT(CUDARuntimeCheck, unsigned, 1, 0)
 
+/// If true, do not generate definition for runtime type info global objects of
+/// derived types defined in other compilation units.
+ENUM_LOWERINGOPT(SkipExternalRttiDefinition, unsigned, 1, 0)
+
 #undef LOWERINGOPT
 #undef ENUM_LOWERINGOPT
diff --git a/flang/include/flang/Optimizer/CodeGen/CodeGen.h b/flang/include/flang/Optimizer/CodeGen/CodeGen.h
index 0398d0f2..93f07d8 100644
--- a/flang/include/flang/Optimizer/CodeGen/CodeGen.h
+++ b/flang/include/flang/Optimizer/CodeGen/CodeGen.h
@@ -39,6 +39,9 @@ struct FIRToLLVMPassOptions {
   // that such programs would crash at runtime if the derived type descriptors
   // are required by the runtime, so this is only an option to help debugging.
   bool ignoreMissingTypeDescriptors = false;
+  // Similar to ignoreMissingTypeDescriptors, but generate external declaration
+  // for the missing type descriptor globals instead.
+  bool skipExternalRttiDefinition = false;
 
   // Generate TBAA information for FIR types and memory accessing operations.
   bool applyTBAA = false;
diff --git a/flang/include/flang/Optimizer/Passes/CommandLineOpts.h b/flang/include/flang/Optimizer/Passes/CommandLineOpts.h
index 1cfaf28..76ac9d0 100644
--- a/flang/include/flang/Optimizer/Passes/CommandLineOpts.h
+++ b/flang/include/flang/Optimizer/Passes/CommandLineOpts.h
@@ -32,6 +32,19 @@ extern llvm::cl::opt<std::size_t> arrayStackAllocationThreshold;
 /// generated by the frontend.
 extern llvm::cl::opt<bool> ignoreMissingTypeDescriptors;
 
+/// Shared option in tools to only generate rtti static object definitions for
+/// derived types defined in the current compilation unit. Derived type
+/// descriptor object for types defined in other objects will only be declared
+/// as external. This also changes the linkage of rtti objects defined in the
+/// current compilation unit from linkonce_odr to external so that unused rtti
+/// objects are retained and can be accessed from other compilation units. This
+/// is an experimental option to explore compilation speed improvements and is
+/// an ABI breaking change because of the linkage change.
+/// It will also require linking against module file objects of modules defining
+/// only types (even for trivial types without type bound procedures, which
+/// differs from most compilers).
+extern llvm::cl::opt<bool> skipExternalRttiDefinition;
+
 /// Default optimization level used to create Flang pass pipeline is O0.
 extern llvm::OptimizationLevel defaultOptLevel;
 
diff --git a/flang/include/flang/Optimizer/Support/Utils.h b/flang/include/flang/Optimizer/Support/Utils.h
index ec73af6..83c936b 100644
--- a/flang/include/flang/Optimizer/Support/Utils.h
+++ b/flang/include/flang/Optimizer/Support/Utils.h
@@ -35,32 +35,6 @@ inline std::int64_t toInt(mlir::arith::ConstantOp cop) {
       .getSExtValue();
 }
 
-// Reconstruct binding tables for dynamic dispatch.
-using BindingTable = llvm::DenseMap<llvm::StringRef, unsigned>;
-using BindingTables = llvm::DenseMap<llvm::StringRef, BindingTable>;
-
-inline void buildBindingTables(BindingTables &bindingTables,
-                               mlir::ModuleOp mod) {
-
-  // The binding tables are defined in FIR after lowering inside fir.type_info
-  // operations. Go through each binding tables and store the procedure name and
-  // binding index for later use by the fir.dispatch conversion pattern.
-  for (auto typeInfo : mod.getOps<fir::TypeInfoOp>()) {
-    unsigned bindingIdx = 0;
-    BindingTable bindings;
-    if (typeInfo.getDispatchTable().empty()) {
-      bindingTables[typeInfo.getSymName()] = bindings;
-      continue;
-    }
-    for (auto dtEntry :
-         typeInfo.getDispatchTable().front().getOps<fir::DTEntryOp>()) {
-      bindings[dtEntry.getMethod()] = bindingIdx;
-      ++bindingIdx;
-    }
-    bindingTables[typeInfo.getSymName()] = bindings;
-  }
-}
-
 // Translate front-end KINDs for use in the IR and code gen.
 inline std::vector<fir::KindTy>
 fromDefaultKinds(const Fortran::common::IntrinsicTypeDefaultKinds &defKinds) {
diff --git a/flang/include/flang/Semantics/runtime-type-info.h b/flang/include/flang/Semantics/runtime-type-info.h
index e90d3ae..6c5a061 100644
--- a/flang/include/flang/Semantics/runtime-type-info.h
+++ b/flang/include/flang/Semantics/runtime-type-info.h
@@ -38,6 +38,10 @@ RuntimeDerivedTypeTables BuildRuntimeDerivedTypeTables(SemanticsContext &);
 /// to describe other derived types at runtime in flang descriptor.
 constexpr char typeInfoBuiltinModule[]{"__fortran_type_info"};
 
+/// Name of the builtin derived type in __fortran_type_inf that is used for
+/// derived type descriptors.
+constexpr char typeDescriptorTypeName[]{"derivedtype"};
+
 /// Name of the bindings descriptor component in the DerivedType type of the
 /// __Fortran_type_info module
 constexpr char bindingDescCompName[]{"binding"};
diff --git a/flang/lib/Evaluate/tools.cpp b/flang/lib/Evaluate/tools.cpp
index 6883856..fcacdb93 100644
--- a/flang/lib/Evaluate/tools.cpp
+++ b/flang/lib/Evaluate/tools.cpp
@@ -2334,6 +2334,11 @@ bool IsBuiltinCPtr(const Symbol &symbol) {
   return false;
 }
 
+bool IsFromBuiltinModule(const Symbol &symbol) {
+  const Scope &scope{symbol.GetUltimate().owner()};
+  return IsSameModule(&scope, scope.context().GetBuiltinsScope());
+}
+
 bool IsIsoCType(const DerivedTypeSpec *derived) {
   return IsBuiltinDerivedType(derived, "c_ptr") ||
       IsBuiltinDerivedType(derived, "c_funptr");
diff --git a/flang/lib/Frontend/CompilerInvocation.cpp b/flang/lib/Frontend/CompilerInvocation.cpp
index 766131f..30d81f3 100644
--- a/flang/lib/Frontend/CompilerInvocation.cpp
+++ b/flang/lib/Frontend/CompilerInvocation.cpp
@@ -14,6 +14,7 @@
 #include "flang/Frontend/CodeGenOptions.h"
 #include "flang/Frontend/PreprocessorOptions.h"
 #include "flang/Frontend/TargetOptions.h"
+#include "flang/Optimizer/Passes/CommandLineOpts.h"
 #include "flang/Semantics/semantics.h"
 #include "flang/Support/Fortran-features.h"
 #include "flang/Support/OpenMP-features.h"
@@ -1792,6 +1793,7 @@ void CompilerInvocation::setLoweringOptions() {
   // Lower TRANSPOSE as a runtime call under -O0.
   loweringOpts.setOptimizeTranspose(codegenOpts.OptimizationLevel > 0);
   loweringOpts.setUnderscoring(codegenOpts.Underscoring);
+  loweringOpts.setSkipExternalRttiDefinition(skipExternalRttiDefinition);
 
   const Fortran::common::LangOptions &langOptions = getLangOpts();
   loweringOpts.setIntegerWrapAround(langOptions.getSignedOverflowBehavior() ==
diff --git a/flang/lib/Lower/Bridge.cpp b/flang/lib/Lower/Bridge.cpp
index 7b640dd..ff35840a 100644
--- a/flang/lib/Lower/Bridge.cpp
+++ b/flang/lib/Lower/Bridge.cpp
@@ -262,6 +262,7 @@ public:
   }
 
   void createTypeInfo(Fortran::lower::AbstractConverter &converter) {
+    createTypeInfoForTypeDescriptorBuiltinType(converter);
     while (!registeredTypeInfoA.empty()) {
       currentTypeInfoStack = &registeredTypeInfoB;
       for (const TypeInfo &info : registeredTypeInfoA)
@@ -277,10 +278,22 @@ public:
 private:
   void createTypeInfoOpAndGlobal(Fortran::lower::AbstractConverter &converter,
                                  const TypeInfo &info) {
-    Fortran::lower::createRuntimeTypeInfoGlobal(converter, info.symbol.get());
+    if (!converter.getLoweringOptions().getSkipExternalRttiDefinition())
+      Fortran::lower::createRuntimeTypeInfoGlobal(converter, info.symbol.get());
     createTypeInfoOp(converter, info);
   }
 
+  void createTypeInfoForTypeDescriptorBuiltinType(
+      Fortran::lower::AbstractConverter &converter) {
+    if (registeredTypeInfoA.empty())
+      return;
+    auto builtinTypeInfoType = llvm::cast<fir::RecordType>(
+        converter.genType(registeredTypeInfoA[0].symbol.get()));
+    converter.getFirOpBuilder().createTypeInfoOp(
+        registeredTypeInfoA[0].loc, builtinTypeInfoType,
+        /*parentType=*/fir::RecordType{});
+  }
+
   void createTypeInfoOp(Fortran::lower::AbstractConverter &converter,
                         const TypeInfo &info) {
     fir::RecordType parentType{};
diff --git a/flang/lib/Lower/ConvertVariable.cpp b/flang/lib/Lower/ConvertVariable.cpp
index 49e6ea02..7ab3c43 100644
--- a/flang/lib/Lower/ConvertVariable.cpp
+++ b/flang/lib/Lower/ConvertVariable.cpp
@@ -647,13 +647,19 @@ fir::GlobalOp Fortran::lower::defineGlobal(
 
 /// Return linkage attribute for \p var.
 static mlir::StringAttr
-getLinkageAttribute(fir::FirOpBuilder &builder,
+getLinkageAttribute(Fortran::lower::AbstractConverter &converter,
                     const Fortran::lower::pft::Variable &var) {
+  fir::FirOpBuilder &builder = converter.getFirOpBuilder();
   // Runtime type info for a same derived type is identical in each compilation
   // unit. It desired to avoid having to link against module that only define a
   // type. Therefore the runtime type info is generated everywhere it is needed
-  // with `linkonce_odr` LLVM linkage.
-  if (var.isRuntimeTypeInfoData())
+  // with `linkonce_odr` LLVM linkage (unless the skipExternalRttiDefinition
+  // option is set, in which case one will need to link against objects of
+  // modules defining types). Builtin objects rtti is always generated because
+  // the builtin module is currently not compiled or part of the runtime.
+  if (var.isRuntimeTypeInfoData() &&
+      (!converter.getLoweringOptions().getSkipExternalRttiDefinition() ||
+       Fortran::semantics::IsFromBuiltinModule(var.getSymbol())))
     return builder.createLinkOnceODRLinkage();
   if (var.isModuleOrSubmoduleVariable())
     return {}; // external linkage
@@ -673,7 +679,7 @@ static void instantiateGlobal(Fortran::lower::AbstractConverter &converter,
   fir::FirOpBuilder &builder = converter.getFirOpBuilder();
   std::string globalName = converter.mangleName(sym);
   mlir::Location loc = genLocation(converter, sym);
-  mlir::StringAttr linkage = getLinkageAttribute(builder, var);
+  mlir::StringAttr linkage = getLinkageAttribute(converter, var);
   fir::GlobalOp global;
   if (var.isModuleOrSubmoduleVariable()) {
     // A non-intrinsic module global is defined when lowering the module.
@@ -1265,7 +1271,7 @@ instantiateAggregateStore(Fortran::lower::AbstractConverter &converter,
   if (var.isGlobal()) {
     fir::GlobalOp global;
     auto &aggregate = var.getAggregateStore();
-    mlir::StringAttr linkage = getLinkageAttribute(builder, var);
+    mlir::StringAttr linkage = getLinkageAttribute(converter, var);
     if (var.isModuleOrSubmoduleVariable()) {
       // A module global was or will be defined when lowering the module. Emit
       // only a declaration if the global does not exist at that point.
@@ -2470,8 +2476,7 @@ void Fortran::lower::defineModuleVariable(
     AbstractConverter &converter, const Fortran::lower::pft::Variable &var) {
   // Use empty linkage for module variables, which makes them available
   // for use in another unit.
-  mlir::StringAttr linkage =
-      getLinkageAttribute(converter.getFirOpBuilder(), var);
+  mlir::StringAttr linkage = getLinkageAttribute(converter, var);
   if (!var.isGlobal())
     fir::emitFatalError(converter.getCurrentLocation(),
                         "attempting to lower module variable as local");
@@ -2606,10 +2611,9 @@ void Fortran::lower::createIntrinsicModuleGlobal(
 void Fortran::lower::createRuntimeTypeInfoGlobal(
     Fortran::lower::AbstractConverter &converter,
     const Fortran::semantics::Symbol &typeInfoSym) {
-  fir::FirOpBuilder &builder = converter.getFirOpBuilder();
   std::string globalName = converter.mangleName(typeInfoSym);
   auto var = Fortran::lower::pft::Variable(typeInfoSym, /*global=*/true);
-  mlir::StringAttr linkage = getLinkageAttribute(builder, var);
+  mlir::StringAttr linkage = getLinkageAttribute(converter, var);
   defineGlobal(converter, var, globalName, linkage);
 }
 
diff --git a/flang/lib/Optimizer/CodeGen/CodeGen.cpp b/flang/lib/Optimizer/CodeGen/CodeGen.cpp
index a3de3ae..2b01891 100644
--- a/flang/lib/Optimizer/CodeGen/CodeGen.cpp
+++ b/flang/lib/Optimizer/CodeGen/CodeGen.cpp
@@ -1294,6 +1294,51 @@ genCUFAllocDescriptor(mlir::Location loc,
       .getResult();
 }
 
+/// Get the address of the type descriptor global variable that was created by
+/// lowering for derived type \p recType.
+template <typename ModOpTy>
+static mlir::Value
+getTypeDescriptor(ModOpTy mod, mlir::ConversionPatternRewriter &rewriter,
+                  mlir::Location loc, fir::RecordType recType,
+                  const fir::FIRToLLVMPassOptions &options) {
+  std::string name =
+      options.typeDescriptorsRenamedForAssembly
+          ? fir::NameUniquer::getTypeDescriptorAssemblyName(recType.getName())
+          : fir::NameUniquer::getTypeDescriptorName(recType.getName());
+  mlir::Type llvmPtrTy = ::getLlvmPtrType(mod.getContext());
+  if (auto global = mod.template lookupSymbol<fir::GlobalOp>(name))
+    return rewriter.create<mlir::LLVM::AddressOfOp>(loc, llvmPtrTy,
+                                                    global.getSymName());
+  // The global may have already been translated to LLVM.
+  if (auto global = mod.template lookupSymbol<mlir::LLVM::GlobalOp>(name))
+    return rewriter.create<mlir::LLVM::AddressOfOp>(loc, llvmPtrTy,
+                                                    global.getSymName());
+  // Type info derived types do not have type descriptors since they are the
+  // types defining type descriptors.
+  if (options.ignoreMissingTypeDescriptors ||
+      fir::NameUniquer::belongsToModule(
+          name, Fortran::semantics::typeInfoBuiltinModule))
+    return rewriter.create<mlir::LLVM::ZeroOp>(loc, llvmPtrTy);
+
+  if (!options.skipExternalRttiDefinition)
+    fir::emitFatalError(loc,
+                        "runtime derived type info descriptor was not "
+                        "generated and skipExternalRttiDefinition and "
+                        "ignoreMissingTypeDescriptors options are not set");
+
+  // Rtti for a derived type defined in another compilation unit and for which
+  // rtti was not defined in lowering because of the skipExternalRttiDefinition
+  // option. Generate the object declaration now.
+  auto insertPt = rewriter.saveInsertionPoint();
+  rewriter.setInsertionPoint(mod.getBody(), mod.getBody()->end());
+  mlir::LLVM::GlobalOp global = rewriter.create<mlir::LLVM::GlobalOp>(
+      loc, llvmPtrTy, /*constant=*/true, mlir::LLVM::Linkage::External, name,
+      mlir::Attribute());
+  rewriter.restoreInsertionPoint(insertPt);
+  return rewriter.create<mlir::LLVM::AddressOfOp>(loc, llvmPtrTy,
+                                                  global.getSymName());
+}
+
 /// Common base class for embox to descriptor conversion.
 template <typename OP>
 struct EmboxCommonConversion : public fir::FIROpConversion<OP> {
@@ -1406,36 +1451,6 @@ struct EmboxCommonConversion : public fir::FIROpConversion<OP> {
                        stride);
   }
 
-  /// Get the address of the type descriptor global variable that was created by
-  /// lowering for derived type \p recType.
-  template <typename ModOpTy>
-  mlir::Value
-  getTypeDescriptor(ModOpTy mod, mlir::ConversionPatternRewriter &rewriter,
-                    mlir::Location loc, fir::RecordType recType) const {
-    std::string name =
-        this->options.typeDescriptorsRenamedForAssembly
-            ? fir::NameUniquer::getTypeDescriptorAssemblyName(recType.getName())
-            : fir::NameUniquer::getTypeDescriptorName(recType.getName());
-    mlir::Type llvmPtrTy = ::getLlvmPtrType(mod.getContext());
-    if (auto global = mod.template lookupSymbol<fir::GlobalOp>(name)) {
-      return rewriter.create<mlir::LLVM::AddressOfOp>(loc, llvmPtrTy,
-                                                      global.getSymName());
-    }
-    if (auto global = mod.template lookupSymbol<mlir::LLVM::GlobalOp>(name)) {
-      // The global may have already been translated to LLVM.
-      return rewriter.create<mlir::LLVM::AddressOfOp>(loc, llvmPtrTy,
-                                                      global.getSymName());
-    }
-    // Type info derived types do not have type descriptors since they are the
-    // types defining type descriptors.
-    if (!this->options.ignoreMissingTypeDescriptors &&
-        !fir::NameUniquer::belongsToModule(
-            name, Fortran::semantics::typeInfoBuiltinModule))
-      fir::emitFatalError(
-          loc, "runtime derived type info descriptor was not generated");
-    return rewriter.create<mlir::LLVM::ZeroOp>(loc, llvmPtrTy);
-  }
-
   template <typename ModOpTy>
   mlir::Value populateDescriptor(mlir::Location loc, ModOpTy mod,
                                  fir::BaseBoxType boxTy, mlir::Type inputType,
@@ -1500,7 +1515,8 @@ struct EmboxCommonConversion : public fir::FIROpConversion<OP> {
           mlir::Type innerType = fir::unwrapInnerType(inputType);
           if (innerType && mlir::isa<fir::RecordType>(innerType)) {
             auto recTy = mlir::dyn_cast<fir::RecordType>(innerType);
-            typeDesc = getTypeDescriptor(mod, rewriter, loc, recTy);
+            typeDesc =
+                getTypeDescriptor(mod, rewriter, loc, recTy, this->options);
           } else {
             // Unlimited polymorphic type descriptor with no record type. Set
             // type descriptor address to a clean state.
@@ -1508,8 +1524,8 @@ struct EmboxCommonConversion : public fir::FIROpConversion<OP> {
                 loc, ::getLlvmPtrType(mod.getContext()));
           }
         } else {
-          typeDesc = getTypeDescriptor(mod, rewriter, loc,
-                                       fir::unwrapIfDerived(boxTy));
+          typeDesc = getTypeDescriptor(
+              mod, rewriter, loc, fir::unwrapIfDerived(boxTy), this->options);
         }
       }
       if (typeDesc)
@@ -3021,22 +3037,10 @@ struct TypeDescOpConversion : public fir::FIROpConversion<fir::TypeDescOp> {
     assert(mlir::isa<fir::RecordType>(inTy) && "expecting fir.type");
     auto recordType = mlir::dyn_cast<fir::RecordType>(inTy);
     auto module = typeDescOp.getOperation()->getParentOfType<mlir::ModuleOp>();
-    std::string typeDescName =
-        this->options.typeDescriptorsRenamedForAssembly
-            ? fir::NameUniquer::getTypeDescriptorAssemblyName(
-                  recordType.getName())
-            : fir::NameUniquer::getTypeDescriptorName(recordType.getName());
-    auto llvmPtrTy = ::getLlvmPtrType(typeDescOp.getContext());
-    if (auto global = module.lookupSymbol<mlir::LLVM::GlobalOp>(typeDescName)) {
-      rewriter.replaceOpWithNewOp<mlir::LLVM::AddressOfOp>(
-          typeDescOp, llvmPtrTy, global.getSymName());
-      return mlir::success();
-    } else if (auto global = module.lookupSymbol<fir::GlobalOp>(typeDescName)) {
-      rewriter.replaceOpWithNewOp<mlir::LLVM::AddressOfOp>(
-          typeDescOp, llvmPtrTy, global.getSymName());
-      return mlir::success();
-    }
-    return mlir::failure();
+    mlir::Value typeDesc = getTypeDescriptor(
+        module, rewriter, typeDescOp.getLoc(), recordType, this->options);
+    rewriter.replaceOp(typeDescOp, typeDesc);
+    return mlir::success();
   }
 };
 
diff --git a/flang/lib/Optimizer/Passes/CommandLineOpts.cpp b/flang/lib/Optimizer/Passes/CommandLineOpts.cpp
index f95a280..0142375 100644
--- a/flang/lib/Optimizer/Passes/CommandLineOpts.cpp
+++ b/flang/lib/Optimizer/Passes/CommandLineOpts.cpp
@@ -39,6 +39,12 @@ cl::opt<bool> ignoreMissingTypeDescriptors(
              "translating FIR to LLVM"),
     cl::init(false), cl::Hidden);
 
+cl::opt<bool> skipExternalRttiDefinition(
+    "skip-external-rtti-definition", llvm::cl::init(false),
+    llvm::cl::desc("do not define rtti static objects for types belonging to "
+                   "other compilation units"),
+    cl::Hidden);
+
 OptimizationLevel defaultOptLevel{OptimizationLevel::O0};
 
 codegenoptions::DebugInfoKind noDebugInfo{codegenoptions::NoDebugInfo};
diff --git a/flang/lib/Optimizer/Passes/Pipelines.cpp b/flang/lib/Optimizer/Passes/Pipelines.cpp
index 70f57bd..42d9e7b 100644
--- a/flang/lib/Optimizer/Passes/Pipelines.cpp
+++ b/flang/lib/Optimizer/Passes/Pipelines.cpp
@@ -108,6 +108,7 @@ void addFIRToLLVMPass(mlir::PassManager &pm,
                       const MLIRToLLVMPassPipelineConfig &config) {
   fir::FIRToLLVMPassOptions options;
   options.ignoreMissingTypeDescriptors = ignoreMissingTypeDescriptors;
+  options.skipExternalRttiDefinition = skipExternalRttiDefinition;
   options.applyTBAA = config.AliasAnalysis;
   options.forceUnifiedTBAATree = useOldAliasTags;
   options.typeDescriptorsRenamedForAssembly =
diff --git a/flang/lib/Optimizer/Transforms/PolymorphicOpConversion.cpp b/flang/lib/Optimizer/Transforms/PolymorphicOpConversion.cpp
index f9a4c4d..57eae1f 100644
--- a/flang/lib/Optimizer/Transforms/PolymorphicOpConversion.cpp
+++ b/flang/lib/Optimizer/Transforms/PolymorphicOpConversion.cpp
@@ -16,7 +16,6 @@
 #include "flang/Optimizer/Dialect/Support/KindMapping.h"
 #include "flang/Optimizer/Support/InternalNames.h"
 #include "flang/Optimizer/Support/TypeCode.h"
-#include "flang/Optimizer/Support/Utils.h"
 #include "flang/Optimizer/Transforms/Passes.h"
 #include "flang/Runtime/derived-api.h"
 #include "flang/Semantics/runtime-type-info.h"
@@ -38,6 +37,45 @@ namespace fir {
 using namespace fir;
 using namespace mlir;
 
+// Reconstruct binding tables for dynamic dispatch.
+using BindingTable = llvm::DenseMap<llvm::StringRef, unsigned>;
+using BindingTables = llvm::DenseMap<llvm::StringRef, BindingTable>;
+
+static std::string getTypeDescriptorTypeName() {
+  llvm::SmallVector<llvm::StringRef, 1> modules = {
+      Fortran::semantics::typeInfoBuiltinModule};
+  return fir::NameUniquer::doType(modules, /*proc=*/{}, /*blockId=*/0,
+                                  Fortran::semantics::typeDescriptorTypeName,
+                                  /*kinds=*/{});
+}
+
+static std::optional<mlir::Type>
+buildBindingTables(BindingTables &bindingTables, mlir::ModuleOp mod) {
+
+  std::optional<mlir::Type> typeDescriptorType;
+  std::string typeDescriptorTypeName = getTypeDescriptorTypeName();
+  // The binding tables are defined in FIR after lowering inside fir.type_info
+  // operations. Go through each binding tables and store the procedure name and
+  // binding index for later use by the fir.dispatch conversion pattern.
+  for (auto typeInfo : mod.getOps<fir::TypeInfoOp>()) {
+    if (!typeDescriptorType && typeInfo.getSymName() == typeDescriptorTypeName)
+      typeDescriptorType = typeInfo.getType();
+    unsigned bindingIdx = 0;
+    BindingTable bindings;
+    if (typeInfo.getDispatchTable().empty()) {
+      bindingTables[typeInfo.getSymName()] = bindings;
+      continue;
+    }
+    for (auto dtEntry :
+         typeInfo.getDispatchTable().front().getOps<fir::DTEntryOp>()) {
+      bindings[dtEntry.getMethod()] = bindingIdx;
+      ++bindingIdx;
+    }
+    bindingTables[typeInfo.getSymName()] = bindings;
+  }
+  return typeDescriptorType;
+}
+
 namespace {
 
 /// SelectTypeOp converted to an if-then-else chain
@@ -77,9 +115,10 @@ private:
 struct DispatchOpConv : public OpConversionPattern<fir::DispatchOp> {
   using OpConversionPattern<fir::DispatchOp>::OpConversionPattern;
 
-  DispatchOpConv(mlir::MLIRContext *ctx, const BindingTables &bindingTables)
+  DispatchOpConv(mlir::MLIRContext *ctx, const BindingTables &bindingTables,
+                 std::optional<mlir::Type> typeDescriptorType)
       : mlir::OpConversionPattern<fir::DispatchOp>(ctx),
-        bindingTables(bindingTables) {}
+        bindingTables(bindingTables), typeDescriptorType{typeDescriptorType} {}
 
   llvm::LogicalResult
   matchAndRewrite(fir::DispatchOp dispatch, OpAdaptor adaptor,
@@ -111,13 +150,11 @@ struct DispatchOpConv : public OpConversionPattern<fir::DispatchOp> {
 
     mlir::Value passedObject = dispatch.getObject();
 
-    auto module = dispatch.getOperation()->getParentOfType<mlir::ModuleOp>();
-    Type typeDescTy;
-    std::string typeDescName =
-        NameUniquer::getTypeDescriptorName(recordType.getName());
-    if (auto global = module.lookupSymbol<fir::GlobalOp>(typeDescName)) {
-      typeDescTy = global.getType();
-    }
+    if (!typeDescriptorType)
+      return emitError(loc) << "cannot find " << getTypeDescriptorTypeName()
+                            << " fir.type_info that is required to get the "
+                               "related builtin type and lower fir.dispatch";
+    mlir::Type typeDescTy = *typeDescriptorType;
 
     // clang-format off
     // Before:
@@ -213,6 +250,7 @@ struct DispatchOpConv : public OpConversionPattern<fir::DispatchOp> {
 
 private:
   BindingTables bindingTables;
+  std::optional<mlir::Type> typeDescriptorType;
 };
 
 /// Convert FIR structured control flow ops to CFG ops.
@@ -229,10 +267,11 @@ public:
     mlir::RewritePatternSet patterns(context);
 
     BindingTables bindingTables;
-    buildBindingTables(bindingTables, mod);
+    std::optional<mlir::Type> typeDescriptorType =
+        buildBindingTables(bindingTables, mod);
 
     patterns.insert<SelectTypeConv>(context);
-    patterns.insert<DispatchOpConv>(context, bindingTables);
+    patterns.insert<DispatchOpConv>(context, bindingTables, typeDescriptorType);
     mlir::ConversionTarget target(*context);
     target.addLegalDialect<mlir::affine::AffineDialect,
                            mlir::cf::ControlFlowDialect, FIROpsDialect,
@@ -379,16 +418,11 @@ llvm::LogicalResult SelectTypeConv::genTypeLadderStep(
   } else if (auto a = mlir::dyn_cast<fir::SubclassAttr>(attr)) {
     // Retrieve the type descriptor from the type guard statement record type.
     assert(mlir::isa<fir::RecordType>(a.getType()) && "expect fir.record type");
-    fir::RecordType recTy = mlir::dyn_cast<fir::RecordType>(a.getType());
-    std::string typeDescName =
-        fir::NameUniquer::getTypeDescriptorName(recTy.getName());
-    auto typeDescGlobal = mod.lookupSymbol<fir::GlobalOp>(typeDescName);
-    auto typeDescAddr = rewriter.create<fir::AddrOfOp>(
-        loc, fir::ReferenceType::get(typeDescGlobal.getType()),
-        typeDescGlobal.getSymbol());
-    mlir::Type typeDescTy = ReferenceType::get(rewriter.getNoneType());
+    mlir::Value typeDescAddr =
+        rewriter.create<fir::TypeDescOp>(loc, mlir::TypeAttr::get(a.getType()));
+    mlir::Type refNoneType = ReferenceType::get(rewriter.getNoneType());
     mlir::Value typeDesc =
-        rewriter.create<ConvertOp>(loc, typeDescTy, typeDescAddr);
+        rewriter.create<ConvertOp>(loc, refNoneType, typeDescAddr);
 
     // Prepare the selector descriptor for the runtime call.
     mlir::Type descNoneTy = fir::BoxType::get(rewriter.getNoneType());
@@ -406,7 +440,7 @@ llvm::LogicalResult SelectTypeConv::genTypeLadderStep(
                                mlir::UnitAttr::get(rewriter.getContext()));
       callee =
           fir::createFuncOp(rewriter.getUnknownLoc(), mod, fctName,
-                            rewriter.getFunctionType({descNoneTy, typeDescTy},
+                            rewriter.getFunctionType({descNoneTy, refNoneType},
                                                      rewriter.getI1Type()),
                             {runtimeAttr});
     }
@@ -435,20 +469,11 @@ SelectTypeConv::genTypeDescCompare(mlir::Location loc, mlir::Value selector,
                                    mlir::Type ty, mlir::ModuleOp mod,
                                    mlir::PatternRewriter &rewriter) const {
   assert(mlir::isa<fir::RecordType>(ty) && "expect fir.record type");
-  fir::RecordType recTy = mlir::dyn_cast<fir::RecordType>(ty);
-  std::string typeDescName =
-      fir::NameUniquer::getTypeDescriptorName(recTy.getName());
-  auto typeDescGlobal = mod.lookupSymbol<fir::GlobalOp>(typeDescName);
-  if (!typeDescGlobal)
-    return {};
-  auto typeDescAddr = rewriter.create<fir::AddrOfOp>(
-      loc, fir::ReferenceType::get(typeDescGlobal.getType()),
-      typeDescGlobal.getSymbol());
+  mlir::Value typeDescAddr =
+      rewriter.create<fir::TypeDescOp>(loc, mlir::TypeAttr::get(ty));
+  mlir::Value selectorTdescAddr = rewriter.create<fir::BoxTypeDescOp>(
+      loc, typeDescAddr.getType(), selector);
   auto intPtrTy = rewriter.getIndexType();
-  mlir::Type tdescType =
-      fir::TypeDescType::get(mlir::NoneType::get(rewriter.getContext()));
-  mlir::Value selectorTdescAddr =
-      rewriter.create<fir::BoxTypeDescOp>(loc, tdescType, selector);
   auto typeDescInt =
       rewriter.create<fir::ConvertOp>(loc, intPtrTy, typeDescAddr);
   auto selectorTdescInt =
diff --git a/flang/test/Integration/skip-external-rtti-definition.F90 b/flang/test/Integration/skip-external-rtti-definition.F90
new file mode 100644
index 0000000..f86cc06
--- /dev/null
+++ b/flang/test/Integration/skip-external-rtti-definition.F90
@@ -0,0 +1,47 @@
+! Test -skip-external-rtti-definition option
+
+!RUN: rm -rf %t && mkdir -p %t
+!RUN: %flang_fc1 -fsyntax-only -DSTEP=1 -J%t %s
+!RUN: %flang_fc1 -emit-llvm -J%t  %s -o - | FileCheck %s -check-prefix=LINKONCE
+!RUN: %flang_fc1 -emit-llvm -J%t -mllvm -skip-external-rtti-definition %s -o - | FileCheck %s -check-prefix=EXTERNAL
+
+#if STEP == 1
+module module_external_type_definition
+ type t1
+ end type
+end module
+#else
+
+module module_same_unit_type_definition
+ type t2
+ end type
+end module
+
+subroutine test
+  use module_external_type_definition
+  use module_same_unit_type_definition
+  interface
+  subroutine needs_descriptor(x)
+    class(*) :: x
+  end subroutine
+  end interface
+  type(t1) :: x1
+  type(t2) :: x2
+  call needs_descriptor(x1)
+  call needs_descriptor(x2)
+end subroutine
+
+#endif
+
+! LINKONCE-DAG: @_QMmodule_external_type_definitionEXnXt1 = linkonce_odr constant [2 x i8] c"t1", comdat
+! LINKONCE-DAG: @_QMmodule_external_type_definitionEXdtXt1 = linkonce_odr constant {{.*}} {
+! LINKONCE-DAG: @_QMmodule_same_unit_type_definitionEXnXt2 = linkonce_odr constant [2 x i8] c"t2", comdat
+! LINKONCE-DAG: @_QMmodule_same_unit_type_definitionEXdtXt2 = linkonce_odr constant {{.*}} {
+
+! EXTERNAL-NOT: @_QMmodule_external_type_definitionEXnXt1
+! EXTERNAL: @_QMmodule_same_unit_type_definitionEXnXt2 = constant [2 x i8] c"t2"
+! EXTERNAL-NOT: @_QMmodule_external_type_definitionEXnXt1
+! EXTERNAL: @_QMmodule_same_unit_type_definitionEXdtXt2 = constant {{.*}} {
+! EXTERNAL-NOT: @_QMmodule_external_type_definitionEXnXt1
+! EXTERNAL: @_QMmodule_external_type_definitionEXdtXt1 =  external constant ptr
+! EXTERNAL-NOT: @_QMmodule_external_type_definitionEXnXt1
diff --git a/flang/test/Lower/select-type-2.f90 b/flang/test/Lower/select-type-2.f90
index 7b4cf96..59485c1 100644
--- a/flang/test/Lower/select-type-2.f90
+++ b/flang/test/Lower/select-type-2.f90
@@ -30,8 +30,8 @@ contains
 
 ! CHECK-LABEL: func.func @_QMselect_type_2Pselect_type1(
 ! CHECK-SAME: %[[ARG0:.*]]: !fir.class<!fir.type<_QMselect_type_2Tp1{a:i32,b:i32}>> {fir.bindc_name = "a"}) {
-! CHECK:      %[[TDESC_P3_ADDR:.*]] = fir.address_of(@_QMselect_type_2E.dt.p3) : !fir.ref<!fir.type<{{.*}}>>
-! CHECK:      %[[TDESC_P3_CONV:.*]] = fir.convert %[[TDESC_P3_ADDR]] : (!fir.ref<!fir.type<{{.*}}>>) -> !fir.ref<none>
+! CHECK:      %[[TDESC_P3_ADDR:.*]] = fir.type_desc !fir.type<_QMselect_type_2Tp3
+! CHECK:      %[[TDESC_P3_CONV:.*]] = fir.convert %[[TDESC_P3_ADDR]] : (!fir.tdesc{{.*}}>) -> !fir.ref<none>
 ! CHECK:      %[[BOX_NONE:.*]] = fir.convert %[[ARG0]] : (!fir.class<!fir.type<_QMselect_type_2Tp1{a:i32,b:i32}>>) -> !fir.box<none>
 ! CHECK:      %[[CLASS_IS_CMP:.*]] = fir.call @_FortranAClassIs(%[[BOX_NONE]], %[[TDESC_P3_CONV]]) : (!fir.box<none>, !fir.ref<none>) -> i1
 ! CHECK:      cf.cond_br %[[CLASS_IS_CMP]], ^[[CLASS_IS_P3_BLK:.*]], ^[[NOT_CLASS_IS_P3_BLK:.*]]
@@ -40,8 +40,8 @@ contains
 ! CHECK:    ^bb[[CLASS_IS_P1:[0-9]]]:
 ! CHECK:      cf.br ^bb[[END_SELECT_BLK:[0-9]]]
 ! CHECK:    ^[[NOT_CLASS_IS_P3_BLK]]:
-! CHECK:      %[[TDESC_P1_ADDR:.*]] = fir.address_of(@_QMselect_type_2E.dt.p1) : !fir.ref<!fir.type<{{.*}}>>
-! CHECK:      %[[TDESC_P1_CONV:.*]] = fir.convert %[[TDESC_P1_ADDR]] : (!fir.ref<!fir.type<{{.*}}>>) -> !fir.ref<none>
+! CHECK:      %[[TDESC_P1_ADDR:.*]] = fir.type_desc !fir.type<_QMselect_type_2Tp1
+! CHECK:      %[[TDESC_P1_CONV:.*]] = fir.convert %[[TDESC_P1_ADDR]] : (!fir.tdesc{{.*}}>) -> !fir.ref<none>
 ! CHECK:      %[[BOX_NONE:.*]] = fir.convert %[[ARG0]] : (!fir.class<!fir.type<_QMselect_type_2Tp1{a:i32,b:i32}>>) -> !fir.box<none>
 ! CHECK:      %[[CLASS_IS_CMP:.*]] = fir.call @_FortranAClassIs(%[[BOX_NONE]], %[[TDESC_P1_CONV]]) : (!fir.box<none>, !fir.ref<none>) -> i1
 ! CHECK:      cf.cond_br %[[CLASS_IS_CMP]], ^bb[[CLASS_IS_P1]], ^bb[[NOT_CLASS_IS_P1]]
diff --git a/flang/test/Lower/select-type.f90 b/flang/test/Lower/select-type.f90
index 64dd639..e2ca87a 100644
--- a/flang/test/Lower/select-type.f90
+++ b/flang/test/Lower/select-type.f90
@@ -68,15 +68,15 @@ contains
 
 ! CFG-LABEL: func.func @_QMselect_type_lower_testPselect_type1(
 ! CFG-SAME: %[[ARG0:.*]]: !fir.class<!fir.type<_QMselect_type_lower_testTp1{a:i32,b:i32}>> {fir.bindc_name = "a"}) {
-! CFG:      %[[TDESC_P1_ADDR:.*]] = fir.address_of(@_QMselect_type_lower_testE.dt.p1) : !fir.ref<!fir.type<{{.*}}>>
-! CFG:      %[[BOX_TDESC:.*]] = fir.box_tdesc %[[ARG0]] : (!fir.class<!fir.type<_QMselect_type_lower_testTp1{a:i32,b:i32}>>) -> !fir.tdesc<none>
-! CFG:      %[[TDESC_P1_CONV:.*]] = fir.convert %[[TDESC_P1_ADDR]] : (!fir.ref<!fir.type<{{.*}}>>) -> index
-! CFG:      %[[BOX_TDESC_CONV:.*]] = fir.convert %[[BOX_TDESC]] : (!fir.tdesc<none>) -> index
+! CFG:      %[[TDESC_P1_ADDR:.*]] = fir.type_desc !fir.type<_QMselect_type_lower_testTp1{a:i32,b:i32}>
+! CFG:      %[[BOX_TDESC:.*]] = fir.box_tdesc %[[ARG0]] : (!fir.class<!fir.type<_QMselect_type_lower_testTp1{a:i32,b:i32}>>) -> !fir.tdesc<!fir.type<_QMselect_type_lower_testTp1{a:i32,b:i32}>>
+! CFG:      %[[TDESC_P1_CONV:.*]] = fir.convert %[[TDESC_P1_ADDR]] : (!fir.tdesc<!fir.type<_QMselect_type_lower_testTp1{a:i32,b:i32}>>) -> index
+! CFG:      %[[BOX_TDESC_CONV:.*]] = fir.convert %[[BOX_TDESC]] : (!fir.tdesc<!fir.type<_QMselect_type_lower_testTp1{a:i32,b:i32}>>) -> index
 ! CFG:      %[[TDESC_CMP:.*]] = arith.cmpi eq, %[[TDESC_P1_CONV]], %[[BOX_TDESC_CONV]] : index
 ! CFG:      cf.cond_br %[[TDESC_CMP]], ^[[TYPE_IS_P1_BLK:.*]], ^[[NOT_TYPE_IS_P1_BLK:.*]]
 ! CFG:    ^[[NOT_TYPE_IS_P1_BLK]]:
-! CFG:      %[[TDESC_P2_ADDR:.*]] = fir.address_of(@_QMselect_type_lower_testE.dt.p2) : !fir.ref<!fir.type<{{.*}}>>
-! CFG:      %[[TDESC_P2_CONV:.*]] = fir.convert %[[TDESC_P2_ADDR]] : (!fir.ref<!fir.type<{{.*}}>>) -> !fir.ref<none>
+! CFG:      %[[TDESC_P2_ADDR:.*]] = fir.type_desc !fir.type<_QMselect_type_lower_testTp2{{.*}}
+! CFG:      %[[TDESC_P2_CONV:.*]] = fir.convert %[[TDESC_P2_ADDR]] : (!fir.tdesc<{{.*}}>) -> !fir.ref<none>
 ! CFG:      %[[BOX_NONE:.*]] = fir.convert %[[ARG0]] : (!fir.class<!fir.type<_QMselect_type_lower_testTp1{a:i32,b:i32}>>) -> !fir.box<none>
 ! CFG:      %[[CLASS_IS:.*]] = fir.call @_FortranAClassIs(%[[BOX_NONE]], %[[TDESC_P2_CONV]]) : (!fir.box<none>, !fir.ref<none>) -> i1
 ! CFG:      cf.cond_br %[[CLASS_IS]], ^bb[[CLASS_IS_P2_BLK:.*]], ^[[NOT_CLASS_IS_P2_BLK:.*]]
@@ -87,8 +87,8 @@ contains
 ! CFG:    ^bb[[CLASS_IS_P1_BLK:[0-9]]]:
 ! CFG:      cf.br ^[[END_SELECT_BLK:.*]]
 ! CFG:    ^[[NOT_CLASS_IS_P2_BLK]]:
-! CFG:      %[[TDESC_P1_ADDR:.*]] = fir.address_of(@_QMselect_type_lower_testE.dt.p1) : !fir.ref<!fir.type<{{.*}}>>
-! CFG:      %[[TDESC_P1_CONV:.*]] = fir.convert %[[TDESC_P1_ADDR]] : (!fir.ref<!fir.type<{{.*}}>>) -> !fir.ref<none>
+! CFG:      %[[TDESC_P1_ADDR:.*]] = fir.type_desc !fir.type<_QMselect_type_lower_testTp1{a:i32,b:i32}>
+! CFG:      %[[TDESC_P1_CONV:.*]] = fir.convert %[[TDESC_P1_ADDR]] : (!fir.tdesc<{{.*}}>) -> !fir.ref<none>
 ! CFG:      %[[BOX_NONE:.*]] = fir.convert %[[ARG0]] : (!fir.class<!fir.type<_QMselect_type_lower_testTp1{a:i32,b:i32}>>) -> !fir.box<none>
 ! CFG:      %[[CLASS_IS:.*]] = fir.call @_FortranAClassIs(%[[BOX_NONE]], %[[TDESC_P1_CONV]]) : (!fir.box<none>, !fir.ref<none>) -> i1
 ! CFG:      cf.cond_br %[[CLASS_IS]], ^bb[[CLASS_IS_P1_BLK]], ^bb[[NOT_CLASS_IS_P1_BLK]]
@@ -126,15 +126,15 @@ contains
 ! CFG:     %[[GET_CLASS:.*]] = fir.call @_QMselect_type_lower_testPget_class() {{.*}} : () -> !fir.class<!fir.ptr<!fir.type<_QMselect_type_lower_testTp1{a:i32,b:i32}>>>
 ! CFG:     fir.save_result %[[GET_CLASS]] to %[[CLASS_ALLOCA]] : !fir.class<!fir.ptr<!fir.type<_QMselect_type_lower_testTp1{a:i32,b:i32}>>>, !fir.ref<!fir.class<!fir.ptr<!fir.type<_QMselect_type_lower_testTp1{a:i32,b:i32}>>>>
 ! CFG:     %[[LOAD_CLASS:.*]] = fir.load %[[CLASS_ALLOCA]] : !fir.ref<!fir.class<!fir.ptr<!fir.type<_QMselect_type_lower_testTp1{a:i32,b:i32}>>>>
-! CFG:     %[[TDESC_P1_ADDR:.*]] = fir.address_of(@_QMselect_type_lower_testE.dt.p1) : !fir.ref<!fir.type<{{.*}}>>
-! CFG:     %[[CLASS_TDESC:.*]] = fir.box_tdesc %[[LOAD_CLASS]] : (!fir.class<!fir.ptr<!fir.type<_QMselect_type_lower_testTp1{a:i32,b:i32}>>>) -> !fir.tdesc<none>
-! CFG:     %[[TDESC_P1_CONV:.*]] = fir.convert %[[TDESC_P1_ADDR]] : (!fir.ref<!fir.type<{{.*}}>>) -> index
-! CFG:     %[[BOX_TDESC_CONV:.*]] = fir.convert %[[CLASS_TDESC]] : (!fir.tdesc<none>) -> index
+! CFG:     %[[TDESC_P1_ADDR:.*]] = fir.type_desc !fir.type<_QMselect_type_lower_testTp1{a:i32,b:i32}>
+! CFG:     %[[CLASS_TDESC:.*]] = fir.box_tdesc %[[LOAD_CLASS]] : (!fir.class<!fir.ptr<!fir.type<_QMselect_type_lower_testTp1{a:i32,b:i32}>>>) -> !fir.tdesc<{{.*}}>
+! CFG:     %[[TDESC_P1_CONV:.*]] = fir.convert %[[TDESC_P1_ADDR]] : (!fir.tdesc{{.*}}>) -> index
+! CFG:     %[[BOX_TDESC_CONV:.*]] = fir.convert %[[CLASS_TDESC]] : (!fir.tdesc<{{.*}}>) -> index
 ! CFG:     %[[TDESC_CMP:.*]] = arith.cmpi eq, %[[TDESC_P1_CONV]], %[[BOX_TDESC_CONV]] : index
 ! CFG:     cf.cond_br %[[TDESC_CMP]], ^[[TYPE_IS_P1_BLK:.*]], ^[[NOT_TYPE_IS_P1_BLK:.*]]
 ! CFG:   ^[[NOT_TYPE_IS_P1_BLK]]:
-! CFG:     %[[TDESC_P1_ADDR:.*]] = fir.address_of(@_QMselect_type_lower_testE.dt.p1) : !fir.ref<!fir.type<{{.*}}>>
-! CFG:     %[[TDESC_P1_CONV:.*]] = fir.convert %[[TDESC_P1_ADDR]] : (!fir.ref<!fir.type<{{.*}}>>) -> !fir.ref<none>
+! CFG:     %[[TDESC_P1_ADDR:.*]] = fir.type_desc !fir.type<_QMselect_type_lower_testTp1{a:i32,b:i32}>
+! CFG:     %[[TDESC_P1_CONV:.*]] = fir.convert %[[TDESC_P1_ADDR]] : (!fir.tdesc{{.*}}>) -> !fir.ref<none>
 ! CFG:     %[[BOX_NONE:.*]] = fir.convert %[[LOAD_CLASS]] : (!fir.class<!fir.ptr<!fir.type<_QMselect_type_lower_testTp1{a:i32,b:i32}>>>) -> !fir.box<none>
 ! CFG:     %[[CLASS_IS:.*]] = fir.call @_FortranAClassIs(%[[BOX_NONE]], %[[TDESC_P1_CONV]]) : (!fir.box<none>, !fir.ref<none>) -> i1
 ! CFG:     cf.cond_br %[[CLASS_IS]], ^[[CLASS_IS_BLK:.*]], ^[[NOT_CLASS_IS_BLK:.*]]
@@ -176,15 +176,15 @@ contains
 ! CFG-LABEL: func.func @_QMselect_type_lower_testPselect_type3(
 ! CFG-SAME: %[[ARG0:.*]]: !fir.ref<!fir.class<!fir.ptr<!fir.array<?x!fir.type<_QMselect_type_lower_testTp1{a:i32,b:i32}>>>>> {fir.bindc_name = "a"}) {
 ! CFG:      %[[SELECTOR:.*]] = fir.embox %{{.*}} source_box %{{.*}} : (!fir.ref<!fir.type<_QMselect_type_lower_testTp1{a:i32,b:i32}>>, !fir.class<{{.*}}>) -> !fir.class<!fir.type<_QMselect_type_lower_testTp1{a:i32,b:i32}>>
-! CFG:      %[[TDESC_P1_ADDR:.*]] = fir.address_of(@_QMselect_type_lower_testE.dt.p1) : !fir.ref<!fir.type<{{.*}}>>
-! CFG:      %[[SELECTOR_TDESC:.*]] = fir.box_tdesc %[[SELECTOR]] : (!fir.class<!fir.type<_QMselect_type_lower_testTp1{a:i32,b:i32}>>) -> !fir.tdesc<none>
-! CFG:      %[[TDESC_P1_CONV:.*]] = fir.convert %[[TDESC_P1_ADDR]] : (!fir.ref<!fir.type<{{.*}}>>) -> index
-! CFG:      %[[TDESC_CONV:.*]] = fir.convert %[[SELECTOR_TDESC]] : (!fir.tdesc<none>) -> index
+! CFG:      %[[TDESC_P1_ADDR:.*]] = fir.type_desc !fir.type<_QMselect_type_lower_testTp1{a:i32,b:i32}>
+! CFG:      %[[SELECTOR_TDESC:.*]] = fir.box_tdesc %[[SELECTOR]] : (!fir.class<!fir.type<_QMselect_type_lower_testTp1{a:i32,b:i32}>>) -> !fir.tdesc<{{.*}}>
+! CFG:      %[[TDESC_P1_CONV:.*]] = fir.convert %[[TDESC_P1_ADDR]] : (!fir.tdesc{{.*}}>) -> index
+! CFG:      %[[TDESC_CONV:.*]] = fir.convert %[[SELECTOR_TDESC]] : (!fir.tdesc<{{.*}}>) -> index
 ! CFG:      %[[TDESC_CMP:.*]] = arith.cmpi eq, %[[TDESC_P1_CONV]], %[[TDESC_CONV]] : index
 ! CFG:      cf.cond_br %[[TDESC_CMP]], ^[[TYPE_IS_P1_BLK:.*]], ^[[NOT_TYPE_IS_P1_BLK:.*]]
 ! CFG:    ^[[NOT_TYPE_IS_P1_BLK]]:
-! CFG:      %[[TDESC_P1_ADDR:.*]] = fir.address_of(@_QMselect_type_lower_testE.dt.p1) : !fir.ref<!fir.type<{{.*}}>>
-! CFG:      %[[TDESC_P1_CONV:.*]] = fir.convert %[[TDESC_P1_ADDR]] : (!fir.ref<!fir.type<{{.*}}>>) -> !fir.ref<none>
+! CFG:      %[[TDESC_P1_ADDR:.*]] = fir.type_desc !fir.type<_QMselect_type_lower_testTp1{a:i32,b:i32}>
+! CFG:      %[[TDESC_P1_CONV:.*]] = fir.convert %[[TDESC_P1_ADDR]] : (!fir.tdesc{{.*}}>) -> !fir.ref<none>
 ! CFG:      %[[BOX_NONE:.*]] = fir.convert %[[SELECTOR]] : (!fir.class<!fir.type<_QMselect_type_lower_testTp1{a:i32,b:i32}>>) -> !fir.box<none>
 ! CFG:      %[[CLASS_IS:.*]] = fir.call @_FortranAClassIs(%[[BOX_NONE]], %[[TDESC_P1_CONV]]) : (!fir.box<none>, !fir.ref<none>) -> i1
 ! CFG:      cf.cond_br %[[CLASS_IS]], ^[[CLASS_IS_BLK:.*]], ^[[NOT_CLASS_IS:.*]]
@@ -222,25 +222,25 @@ contains
 
 ! CFG-LABEL: func.func @_QMselect_type_lower_testPselect_type4(
 ! CFG-SAME: %[[ARG0:.*]]: !fir.class<!fir.type<_QMselect_type_lower_testTp1{a:i32,b:i32}>> {fir.bindc_name = "a"}) {
-! CFG:      %[[TDESC_P3_8_ADDR:.*]] = fir.address_of(@_QMselect_type_lower_testE.dt.p3.8) : !fir.ref<!fir.type<{{.*}}>>
-! CFG:      %[[BOX_TDESC:.*]] = fir.box_tdesc %[[ARG0]] : (!fir.class<!fir.type<_QMselect_type_lower_testTp1{a:i32,b:i32}>>) -> !fir.tdesc<none>
-! CFG:      %[[TDESC_P3_8_CONV:.*]] = fir.convert %[[TDESC_P3_8_ADDR]] : (!fir.ref<!fir.type<{{.*}}>>) -> index
-! CFG:      %[[BOX_TDESC_CONV:.*]] = fir.convert %[[BOX_TDESC]] : (!fir.tdesc<none>) -> index
+! CFG:      %[[TDESC_P3_8_ADDR:.*]] = fir.type_desc !fir.type<_QMselect_type_lower_testTp3K8
+! CFG:      %[[BOX_TDESC:.*]] = fir.box_tdesc %[[ARG0]] : (!fir.class<!fir.type<_QMselect_type_lower_testTp1{a:i32,b:i32}>>) -> !fir.tdesc<{{.*}}>
+! CFG:      %[[TDESC_P3_8_CONV:.*]] = fir.convert %[[TDESC_P3_8_ADDR]] : (!fir.tdesc{{.*}}>) -> index
+! CFG:      %[[BOX_TDESC_CONV:.*]] = fir.convert %[[BOX_TDESC]] : (!fir.tdesc<{{.*}}>) -> index
 ! CFG:      %[[TDESC_CMP:.*]] = arith.cmpi eq, %[[TDESC_P3_8_CONV]], %[[BOX_TDESC_CONV]] : index
 ! CFG:      cf.cond_br %[[TDESC_CMP]], ^[[P3_8_BLK:.*]], ^[[NOT_P3_8_BLK:.*]]
 ! CFG:    ^[[NOT_P3_8_BLK]]:
-! CFG:      %[[TDESC_P3_4_ADDR:.*]] = fir.address_of(@_QMselect_type_lower_testE.dt.p3.4) : !fir.ref<!fir.type<{{.*}}>>
-! CFG:      %[[BOX_TDESC:.*]] = fir.box_tdesc %[[ARG0]] : (!fir.class<!fir.type<_QMselect_type_lower_testTp1{a:i32,b:i32}>>) -> !fir.tdesc<none>
-! CFG:      %[[TDESC_P3_4_CONV:.*]] = fir.convert %[[TDESC_P3_4_ADDR]] : (!fir.ref<!fir.type<{{.*}}>>) -> index
-! CFG:      %[[BOX_TDESC_CONV:.*]] = fir.convert %[[BOX_TDESC]] : (!fir.tdesc<none>) -> index
+! CFG:      %[[TDESC_P3_4_ADDR:.*]] = fir.type_desc !fir.type<_QMselect_type_lower_testTp3K4
+! CFG:      %[[BOX_TDESC:.*]] = fir.box_tdesc %[[ARG0]] : (!fir.class<!fir.type<_QMselect_type_lower_testTp1{a:i32,b:i32}>>) -> !fir.tdesc<{{.*}}>
+! CFG:      %[[TDESC_P3_4_CONV:.*]] = fir.convert %[[TDESC_P3_4_ADDR]] : (!fir.tdesc{{.*}}>) -> index
+! CFG:      %[[BOX_TDESC_CONV:.*]] = fir.convert %[[BOX_TDESC]] : (!fir.tdesc<{{.*}}>) -> index
 ! CFG:      %[[TDESC_CMP:.*]] = arith.cmpi eq, %[[TDESC_P3_4_CONV]], %[[BOX_TDESC_CONV]] : index
 ! CFG:      cf.cond_br %[[TDESC_CMP]], ^[[P3_4_BLK:.*]], ^[[NOT_P3_4_BLK:.*]]
 ! CFG:    ^[[P3_8_BLK]]:
 ! CFG:      _FortranAioOutputAscii
 ! CFG:      cf.br ^bb[[EXIT_SELECT_BLK:[0-9]]]
 ! CFG:    ^[[NOT_P3_4_BLK]]:
-! CFG:      %[[TDESC_P1_ADDR:.*]] = fir.address_of(@_QMselect_type_lower_testE.dt.p1) : !fir.ref<!fir.type<{{.*}}>>
-! CFG:      %[[TDESC_P1_CONV:.*]] = fir.convert %[[TDESC_P1_ADDR]] : (!fir.ref<!fir.type<{{.*}}>>) -> !fir.ref<none>
+! CFG:      %[[TDESC_P1_ADDR:.*]] = fir.type_desc !fir.type<_QMselect_type_lower_testTp1{a:i32,b:i32}>
+! CFG:      %[[TDESC_P1_CONV:.*]] = fir.convert %[[TDESC_P1_ADDR]] : (!fir.tdesc{{.*}}>) -> !fir.ref<none>
 ! CFG:      %[[BOX_NONE:.*]] = fir.convert %[[ARG0]] : (!fir.class<!fir.type<_QMselect_type_lower_testTp1{a:i32,b:i32}>>) -> !fir.box<none>
 ! CFG:      %[[CLASS_IS:.*]] = fir.call @_FortranAClassIs(%[[BOX_NONE]], %[[TDESC_P1_CONV]]) : (!fir.box<none>, !fir.ref<none>) -> i1
 ! CFG:      cf.cond_br %[[CLASS_IS]], ^[[P1_BLK:.*]], ^[[NOT_P1_BLK:.*]]
@@ -409,8 +409,8 @@ contains
 
 ! CFG-LABEL: func.func @_QMselect_type_lower_testPselect_type7(
 ! CFG-SAME: %[[ARG0:.*]]: !fir.class<none> {fir.bindc_name = "a"}) {
-! CFG:      %[[TDESC_P4_ADDR:.*]] = fir.address_of(@_QMselect_type_lower_testE.dt.p4) : !fir.ref<!fir.type<{{.*}}>>
-! CFG:      %[[TDESC_P4_CONV:.*]] = fir.convert %[[TDESC_P4_ADDR]] : (!fir.ref<!fir.type<{{.*}}>>) -> !fir.ref<none>
+! CFG:      %[[TDESC_P4_ADDR:.*]] = fir.type_desc !fir.type<_QMselect_type_lower_testTp4
+! CFG:      %[[TDESC_P4_CONV:.*]] = fir.convert %[[TDESC_P4_ADDR]] : (!fir.tdesc{{.*}}>) -> !fir.ref<none>
 ! CFG:      %[[BOX_NONE:.*]] = fir.convert %[[ARG0]] : (!fir.class<none>) -> !fir.box<none>
 ! CFG:      %[[CLASS_IS_P4:.*]] = fir.call @_FortranAClassIs(%[[BOX_NONE]], %[[TDESC_P4_CONV]]) : (!fir.box<none>, !fir.ref<none>) -> i1
 ! CFG:      cf.cond_br %[[CLASS_IS_P4]], ^[[CLASS_IS_P4_BLK:.*]], ^[[CLASS_NOT_P4_BLK:.*]]
@@ -419,16 +419,16 @@ contains
 ! CFG:    ^bb[[CLASS_IS_P1_BLK:[0-9]]]:
 ! CFG:      cf.br ^[[EXIT_SELECT_BLK:.*]]
 ! CFG:    ^bb[[CLASS_NOT_P2_BLK:[0-9]]]:
-! CFG:      %[[TDESC_P1_ADDR:.*]] = fir.address_of(@_QMselect_type_lower_testE.dt.p1) : !fir.ref<!fir.type<{{.*}}>>
-! CFG:      %[[TDESC_P1_CONV:.*]] = fir.convert %[[TDESC_P1_ADDR]] : (!fir.ref<!fir.type<{{.*}}>>) -> !fir.ref<none>
+! CFG:      %[[TDESC_P1_ADDR:.*]] = fir.type_desc !fir.type<_QMselect_type_lower_testTp1{a:i32,b:i32}>
+! CFG:      %[[TDESC_P1_CONV:.*]] = fir.convert %[[TDESC_P1_ADDR]] : (!fir.tdesc{{.*}}>) -> !fir.ref<none>
 ! CFG:      %[[BOX_NONE:.*]] = fir.convert %[[ARG0]] : (!fir.class<none>) -> !fir.box<none>
 ! CFG:      %[[CLASS_IS_P1:.*]] = fir.call @_FortranAClassIs(%[[BOX_NONE]], %[[TDESC_P1_CONV]]) : (!fir.box<none>, !fir.ref<none>) -> i1
 ! CFG:      cf.cond_br %[[CLASS_IS_P1]], ^bb[[CLASS_IS_P1_BLK]], ^bb[[CLASS_NOT_P1_BLK]]
 ! CFG:    ^bb[[CLASS_IS_P2_BLK:[0-9]]]:
 ! CFG:      cf.br ^[[EXIT_SELECT_BLK]]
 ! CFG:    ^[[CLASS_NOT_P4_BLK]]:
-! CFG:      %[[TDESC_P2_ADDR:.*]] = fir.address_of(@_QMselect_type_lower_testE.dt.p2) : !fir.ref<!fir.type<{{.*}}>>
-! CFG:      %[[TDESC_P2_CONV:.*]] = fir.convert %[[TDESC_P2_ADDR]] : (!fir.ref<!fir.type<{{.*}}>>) -> !fir.ref<none>
+! CFG:      %[[TDESC_P2_ADDR:.*]] = fir.type_desc !fir.type<_QMselect_type_lower_testTp2
+! CFG:      %[[TDESC_P2_CONV:.*]] = fir.convert %[[TDESC_P2_ADDR]] : (!fir.tdesc{{.*}}>) -> !fir.ref<none>
 ! CFG:      %[[BOX_NONE:.*]] = fir.convert %[[ARG0]] : (!fir.class<none>) -> !fir.box<none>
 ! CFG:      %[[CLASS_IS_P2:.*]] = fir.call @_FortranAClassIs(%[[BOX_NONE]], %[[TDESC_P2_CONV]]) : (!fir.box<none>, !fir.ref<none>) -> i1
 ! CFG:      cf.cond_br %[[CLASS_IS_P2]], ^bb[[CLASS_IS_P2_BLK]], ^bb[[CLASS_NOT_P2_BLK]]
diff --git a/flang/tools/bbc/bbc.cpp b/flang/tools/bbc/bbc.cpp
index 015c866..59372a8 100644
--- a/flang/tools/bbc/bbc.cpp
+++ b/flang/tools/bbc/bbc.cpp
@@ -434,6 +434,7 @@ static llvm::LogicalResult convertFortranSourceToMLIR(
   loweringOptions.setStackRepackArrays(stackRepackArrays);
   loweringOptions.setRepackArrays(repackArrays);
   loweringOptions.setRepackArraysWhole(repackArraysWhole);
+  loweringOptions.setSkipExternalRttiDefinition(skipExternalRttiDefinition);
   if (enableCUDA)
     loweringOptions.setCUDARuntimeCheck(true);
   std::vector<Fortran::lower::EnvironmentDefault> envDefaults = {};