Index: lib/Target/AMDGPU/AMDGPU.h
===================================================================
--- lib/Target/AMDGPU/AMDGPU.h
+++ lib/Target/AMDGPU/AMDGPU.h
@@ -90,7 +90,7 @@
 ModulePass *createAMDGPUOpenCLImageTypeLoweringPass();
 FunctionPass *createAMDGPUAnnotateUniformValues();
 
-FunctionPass* createAMDGPUUnifyMetadataPass();
+ModulePass* createAMDGPUUnifyMetadataPass();
 void initializeAMDGPUUnifyMetadataPass(PassRegistry&);
 extern char &AMDGPUUnifyMetadataID;
 
Index: lib/Target/AMDGPU/AMDGPUUnifyMetadata.cpp
===================================================================
--- lib/Target/AMDGPU/AMDGPUUnifyMetadata.cpp
+++ lib/Target/AMDGPU/AMDGPUUnifyMetadata.cpp
@@ -30,21 +30,14 @@
   }
 
   /// \brief Unify multiple OpenCL metadata due to linking.
-  class AMDGPUUnifyMetadata : public FunctionPass {
+  class AMDGPUUnifyMetadata : public ModulePass {
   public:
     static char ID;
-    explicit AMDGPUUnifyMetadata() : FunctionPass(ID) {};
+    explicit AMDGPUUnifyMetadata() : ModulePass(ID) {};
 
   private:
-    // This should really be a module pass but we have to run it as early
-    // as possible, so given function passes are executed first and
-    // TargetMachine::addEarlyAsPossiblePasses() expects only function passes
-    // it has to be a function pass.
     virtual bool runOnModule(Module &M);
 
-    // \todo: Convert to a module pass.
-    virtual bool runOnFunction(Function &F);
-
     /// \brief Unify version metadata.
     /// \return true if changes are made.
     /// Assume the named metadata has operands each of which is a pair of
@@ -117,7 +110,7 @@
                 "Unify multiple OpenCL metadata due to linking",
                 false, false)
 
-FunctionPass* llvm::createAMDGPUUnifyMetadataPass() {
+ModulePass* llvm::createAMDGPUUnifyMetadataPass() {
   return new AMDGPUUnifyMetadata();
 }
 
@@ -143,7 +136,3 @@
 
   return Changed;
 }
-
-bool AMDGPUUnifyMetadata::runOnFunction(Function &F) {
-  return runOnModule(*F.getParent());
-}
Index: lib/Target/NVPTX/NVPTXTargetMachine.cpp
===================================================================
--- lib/Target/NVPTX/NVPTXTargetMachine.cpp
+++ lib/Target/NVPTX/NVPTXTargetMachine.cpp
@@ -170,15 +170,43 @@
   // Add passes that perform straight-line scalar optimizations.
   void addStraightLineScalarOptimizationPasses();
 };
+
+class NVPTXEarlyAsPossibleFunctionWrapperPass : public ModulePass {
+public:
+  static char ID;
+  NVPTXEarlyAsPossibleFunctionWrapperPass(NVPTXTargetMachine &TM_) :
+    ModulePass(ID), TM(TM_) {}
+
+  bool runOnModule(Module &M) override;
+
+private:
+  NVPTXTargetMachine &TM;
+};
 } // end anonymous namespace
 
 TargetPassConfig *NVPTXTargetMachine::createPassConfig(PassManagerBase &PM) {
   return new NVPTXPassConfig(this, PM);
 }
 
+char NVPTXEarlyAsPossibleFunctionWrapperPass::ID = 0;
+
+bool NVPTXEarlyAsPossibleFunctionWrapperPass::runOnModule(Module &M) {
+  legacy::FunctionPassManager *FPM = new legacy::FunctionPassManager(&M);
+
+  FPM->add(createNVVMReflectPass());
+  FPM->add(createNVVMIntrRangePass(TM.getSubtargetImpl()->getSmVersion()));
+
+  FPM->doInitialization();
+  for (auto &F : M.functions())
+    FPM->run(F);
+  FPM->doFinalization();
+  delete FPM;
+
+  return true;
+}
+
 void NVPTXTargetMachine::addEarlyAsPossiblePasses(PassManagerBase &PM) {
-  PM.add(createNVVMReflectPass());
-  PM.add(createNVVMIntrRangePass(Subtarget.getSmVersion()));
+  PM.add(new NVPTXEarlyAsPossibleFunctionWrapperPass(*this));
 }
 
 TargetIRAnalysis NVPTXTargetMachine::getTargetIRAnalysis() {
Index: lib/Target/NVPTX/NVVMIntrRange.cpp
===================================================================
--- lib/Target/NVPTX/NVVMIntrRange.cpp
+++ lib/Target/NVPTX/NVVMIntrRange.cpp
@@ -55,7 +55,8 @@
 }
 
 FunctionPass *llvm::createNVVMIntrRangePass(unsigned int SmVersion) {
-  return new NVVMIntrRange(SmVersion);
+  return new NVVMIntrRange((NVVMIntrRangeSM.getNumOccurrences() > 0) ?
+                             NVVMIntrRangeSM : SmVersion);
 }
 
 char NVVMIntrRange::ID = 0;
Index: test/CodeGen/AMDGPU/unify-metadata.ll
===================================================================
--- test/CodeGen/AMDGPU/unify-metadata.ll
+++ test/CodeGen/AMDGPU/unify-metadata.ll
@@ -14,10 +14,6 @@
 ; ALL-DAG: ![[USED_EXT_1]] = !{!"cl_khr_fp16"}
 ; ALL-DAG: ![[USED_EXT_2]] = !{!"cl_doubles"}
 
-define void @test() {
-   ret void
-}
-
 !opencl.ocl.version = !{!1, !0, !0, !0}
 !llvm.ident = !{!2, !2, !2, !2, !6}
 !opencl.used.extensions = !{!3, !3, !4, !5}
Index: tools/opt/opt.cpp
===================================================================
--- tools/opt/opt.cpp
+++ tools/opt/opt.cpp
@@ -287,14 +287,6 @@
   Builder.SLPVectorize =
       DisableSLPVectorization ? false : OptLevel > 1 && SizeLevel < 2;
 
-  // Add target-specific passes that need to run as early as possible.
-  if (TM)
-    Builder.addExtension(
-        PassManagerBuilder::EP_EarlyAsPossible,
-        [&](const PassManagerBuilder &, legacy::PassManagerBase &PM) {
-          TM->addEarlyAsPossiblePasses(PM);
-        });
-
   if (Coroutines)
     addCoroutinePassesToExtensionPoints(Builder);
 
@@ -674,6 +666,13 @@
   if (OptLevelO3)
     AddOptimizationPasses(Passes, *FPasses, TM.get(), 3, 0);
 
+  // Add target-specific passes that need to run as early as possible.
+  if (TM) {
+    legacy::PassManager EarlyPasses;
+    TM->addEarlyAsPossiblePasses(EarlyPasses);
+    EarlyPasses.run(*M);
+  }
+
   if (FPasses) {
     FPasses->doInitialization();
     for (Function &F : *M)