Index: clang/lib/Driver/Driver.cpp
===================================================================
--- clang/lib/Driver/Driver.cpp
+++ clang/lib/Driver/Driver.cpp
@@ -2784,11 +2784,9 @@
       }
 
       // By default, we produce an action for each device arch.
-      if (!Relocatable || CurPhase < phases::Backend || CompileDeviceOnly) {
-        for (Action *&A : CudaDeviceActions)
-          A = C.getDriver().ConstructPhaseAction(C, Args, CurPhase, A,
-                                                 AssociatedOffloadKind);
-      }
+      for (Action *&A : CudaDeviceActions)
+        A = C.getDriver().ConstructPhaseAction(C, Args, CurPhase, A,
+                                               AssociatedOffloadKind);
 
       return (CompileDeviceOnly && CurPhase == FinalPhase) ? ABRT_Ignore_Host
                                                            : ABRT_Success;
Index: clang/lib/Driver/ToolChains/AMDGPU.h
===================================================================
--- clang/lib/Driver/ToolChains/AMDGPU.h
+++ clang/lib/Driver/ToolChains/AMDGPU.h
@@ -80,6 +80,11 @@
 
   static bool isWave64(const llvm::opt::ArgList &DriverArgs,
                        llvm::AMDGPU::GPUKind Kind);
+  /// Needed for using lto.
+  bool HasNativeLLVMSupport() const override {
+    return true;
+  }
+
 };
 
 class LLVM_LIBRARY_VISIBILITY ROCMToolChain : public AMDGPUToolChain {
Index: clang/lib/Driver/ToolChains/HIP.h
===================================================================
--- clang/lib/Driver/ToolChains/HIP.h
+++ clang/lib/Driver/ToolChains/HIP.h
@@ -38,34 +38,10 @@
                     const char *LinkingOutput) const override;
 
 private:
-  /// \return llvm-link output file name.
-  const char *constructLLVMLinkCommand(Compilation &C, const JobAction &JA,
-                                       const InputInfoList &Inputs,
-                                       const llvm::opt::ArgList &Args,
-                                       llvm::StringRef SubArchName,
-                                       llvm::StringRef OutputFilePrefix) const;
-
-  /// \return opt output file name.
-  const char *constructOptCommand(Compilation &C, const JobAction &JA,
-                                  const InputInfoList &Inputs,
-                                  const llvm::opt::ArgList &Args,
-                                  llvm::StringRef SubArchName,
-                                  llvm::StringRef OutputFilePrefix,
-                                  const char *InputFileName) const;
-
-  /// \return llc output file name.
-  const char *constructLlcCommand(Compilation &C, const JobAction &JA,
-                                  const InputInfoList &Inputs,
-                                  const llvm::opt::ArgList &Args,
-                                  llvm::StringRef SubArchName,
-                                  llvm::StringRef OutputFilePrefix,
-                                  const char *InputFileName,
-                                  bool OutputIsAsm = false) const;
 
   void constructLldCommand(Compilation &C, const JobAction &JA,
                            const InputInfoList &Inputs, const InputInfo &Output,
-                           const llvm::opt::ArgList &Args,
-                           const char *InputFileName) const;
+                           const llvm::opt::ArgList &Args) const;
 };
 
 } // end namespace AMDGCN
Index: clang/lib/Driver/ToolChains/HIP.cpp
===================================================================
--- clang/lib/Driver/ToolChains/HIP.cpp
+++ clang/lib/Driver/ToolChains/HIP.cpp
@@ -49,152 +49,18 @@
   }
   D.Diag(diag::err_drv_no_such_file) << BCName;
 }
-
-static const char *getOutputFileName(Compilation &C, StringRef Base,
-                                     const char *Postfix,
-                                     const char *Extension) {
-  const char *OutputFileName;
-  if (C.getDriver().isSaveTempsEnabled()) {
-    OutputFileName =
-        C.getArgs().MakeArgString(Base.str() + Postfix + "." + Extension);
-  } else {
-    std::string TmpName =
-        C.getDriver().GetTemporaryPath(Base.str() + Postfix, Extension);
-    OutputFileName = C.addTempFile(C.getArgs().MakeArgString(TmpName));
-  }
-  return OutputFileName;
-}
-
-static void addOptLevelArgs(const llvm::opt::ArgList &Args,
-                            llvm::opt::ArgStringList &CmdArgs,
-                            bool IsLlc = false) {
-  if (Arg *A = Args.getLastArg(options::OPT_O_Group)) {
-    StringRef OOpt = "3";
-    if (A->getOption().matches(options::OPT_O4) ||
-        A->getOption().matches(options::OPT_Ofast))
-      OOpt = "3";
-    else if (A->getOption().matches(options::OPT_O0))
-      OOpt = "0";
-    else if (A->getOption().matches(options::OPT_O)) {
-      // Clang and opt support -Os/-Oz; llc only supports -O0, -O1, -O2 and -O3
-      // so we map -Os/-Oz to -O2.
-      // Only clang supports -Og, and maps it to -O1.
-      // We map anything else to -O2.
-      OOpt = llvm::StringSwitch<const char *>(A->getValue())
-                 .Case("1", "1")
-                 .Case("2", "2")
-                 .Case("3", "3")
-                 .Case("s", IsLlc ? "2" : "s")
-                 .Case("z", IsLlc ? "2" : "z")
-                 .Case("g", "1")
-                 .Default("2");
-    }
-    CmdArgs.push_back(Args.MakeArgString("-O" + OOpt));
-  }
-}
 } // namespace
 
-const char *AMDGCN::Linker::constructLLVMLinkCommand(
-    Compilation &C, const JobAction &JA, const InputInfoList &Inputs,
-    const ArgList &Args, StringRef SubArchName,
-    StringRef OutputFilePrefix) const {
-  ArgStringList CmdArgs;
-  // Add the input bc's created by compile step.
-  for (const auto &II : Inputs)
-    CmdArgs.push_back(II.getFilename());
-
-  // Add an intermediate output file.
-  CmdArgs.push_back("-o");
-  auto OutputFileName = getOutputFileName(C, OutputFilePrefix, "-linked", "bc");
-  CmdArgs.push_back(OutputFileName);
-  const char *Exec =
-      Args.MakeArgString(getToolChain().GetProgramPath("llvm-link"));
-  C.addCommand(std::make_unique<Command>(JA, *this, Exec, CmdArgs, Inputs));
-  return OutputFileName;
-}
-
-const char *AMDGCN::Linker::constructOptCommand(
-    Compilation &C, const JobAction &JA, const InputInfoList &Inputs,
-    const llvm::opt::ArgList &Args, llvm::StringRef SubArchName,
-    llvm::StringRef OutputFilePrefix, const char *InputFileName) const {
-  // Construct opt command.
-  ArgStringList OptArgs;
-  // The input to opt is the output from llvm-link.
-  OptArgs.push_back(InputFileName);
-  // Pass optimization arg to opt.
-  addOptLevelArgs(Args, OptArgs);
-  OptArgs.push_back("-mtriple=amdgcn-amd-amdhsa");
-  OptArgs.push_back(Args.MakeArgString("-mcpu=" + SubArchName));
-
-  for (const Arg *A : Args.filtered(options::OPT_mllvm)) {
-    OptArgs.push_back(A->getValue(0));
-  }
-
-  OptArgs.push_back("-o");
-  auto OutputFileName =
-      getOutputFileName(C, OutputFilePrefix, "-optimized", "bc");
-  OptArgs.push_back(OutputFileName);
-  const char *OptExec =
-      Args.MakeArgString(getToolChain().GetProgramPath("opt"));
-  C.addCommand(std::make_unique<Command>(JA, *this, OptExec, OptArgs, Inputs));
-  return OutputFileName;
-}
-
-const char *AMDGCN::Linker::constructLlcCommand(
-    Compilation &C, const JobAction &JA, const InputInfoList &Inputs,
-    const llvm::opt::ArgList &Args, llvm::StringRef SubArchName,
-    llvm::StringRef OutputFilePrefix, const char *InputFileName,
-    bool OutputIsAsm) const {
-  // Construct llc command.
-  ArgStringList LlcArgs;
-  // The input to llc is the output from opt.
-  LlcArgs.push_back(InputFileName);
-  // Pass optimization arg to llc.
-  addOptLevelArgs(Args, LlcArgs, /*IsLlc=*/true);
-  LlcArgs.push_back("-mtriple=amdgcn-amd-amdhsa");
-  LlcArgs.push_back(Args.MakeArgString("-mcpu=" + SubArchName));
-  LlcArgs.push_back(
-      Args.MakeArgString(Twine("-filetype=") + (OutputIsAsm ? "asm" : "obj")));
-
-  // Extract all the -m options
-  std::vector<llvm::StringRef> Features;
-  handleTargetFeaturesGroup(
-    Args, Features, options::OPT_m_amdgpu_Features_Group);
-
-  // Add features to mattr such as xnack
-  std::string MAttrString = "-mattr=";
-  for(auto OneFeature : Features) {
-    MAttrString.append(Args.MakeArgString(OneFeature));
-    if (OneFeature != Features.back())
-      MAttrString.append(",");
-  }
-  if(!Features.empty())
-    LlcArgs.push_back(Args.MakeArgString(MAttrString));
-
-  for (const Arg *A : Args.filtered(options::OPT_mllvm)) {
-    LlcArgs.push_back(A->getValue(0));
-  }
-
-  // Add output filename
-  LlcArgs.push_back("-o");
-  auto LlcOutputFile =
-      getOutputFileName(C, OutputFilePrefix, "", OutputIsAsm ? "s" : "o");
-  LlcArgs.push_back(LlcOutputFile);
-  const char *Llc = Args.MakeArgString(getToolChain().GetProgramPath("llc"));
-  C.addCommand(std::make_unique<Command>(JA, *this, Llc, LlcArgs, Inputs));
-  return LlcOutputFile;
-}
-
 void AMDGCN::Linker::constructLldCommand(Compilation &C, const JobAction &JA,
                                           const InputInfoList &Inputs,
                                           const InputInfo &Output,
-                                          const llvm::opt::ArgList &Args,
-                                          const char *InputFileName) const {
+                                          const llvm::opt::ArgList &Args) const {
   // Construct lld command.
   // The output from ld.lld is an HSA code object file.
   ArgStringList LldArgs{"-flavor",    "gnu", "--no-undefined",
-                        "-shared",    "-o",  Output.getFilename(),
-                        InputFileName};
+                        "-shared",    "-o",  Output.getFilename()};
+  for (auto Input : Inputs)
+    LldArgs.push_back(Input.getFilename());
   const char *Lld = Args.MakeArgString(getToolChain().GetProgramPath("lld"));
   C.addCommand(std::make_unique<Command>(JA, *this, Lld, LldArgs, Inputs));
 }
@@ -243,33 +109,7 @@
   if (JA.getType() == types::TY_HIP_FATBIN)
     return constructHIPFatbinCommand(C, JA, Output.getFilename(), Inputs, Args, *this);
 
-  assert(Inputs.size());
-  if (Inputs.size() == 1 && Inputs[0].getType() == types::TY_Object)
-    return constructLldCommand(C, JA, Inputs, Output, Args,
-                               Inputs[0].getFilename());
-
-  assert(getToolChain().getTriple().getArch() == llvm::Triple::amdgcn &&
-         "Unsupported target");
-
-  std::string SubArchName = JA.getOffloadingArch();
-  assert(StringRef(SubArchName).startswith("gfx") && "Unsupported sub arch");
-
-  // Prefix for temporary file name.
-  std::string Prefix = llvm::sys::path::stem(Inputs[0].getFilename()).str();
-  if (!C.getDriver().isSaveTempsEnabled())
-    Prefix += "-" + SubArchName;
-
-  // Each command outputs different files.
-  const char *LLVMLinkCommand =
-      constructLLVMLinkCommand(C, JA, Inputs, Args, SubArchName, Prefix);
-  const char *OptCommand = constructOptCommand(C, JA, Inputs, Args, SubArchName,
-                                               Prefix, LLVMLinkCommand);
-  if (C.getDriver().isSaveTempsEnabled())
-    constructLlcCommand(C, JA, Inputs, Args, SubArchName, Prefix, OptCommand,
-                        /*OutputIsAsm=*/true);
-  const char *LlcCommand =
-      constructLlcCommand(C, JA, Inputs, Args, SubArchName, Prefix, OptCommand);
-  constructLldCommand(C, JA, Inputs, Output, Args, LlcCommand);
+  return constructLldCommand(C, JA, Inputs, Output, Args);
 }
 
 HIPToolChain::HIPToolChain(const Driver &D, const llvm::Triple &Triple,
@@ -286,7 +126,7 @@
     Action::OffloadKind DeviceOffloadingKind) const {
   HostTC.addClangTargetOptions(DriverArgs, CC1Args, DeviceOffloadingKind);
 
-  StringRef GpuArch = DriverArgs.getLastArgValue(options::OPT_march_EQ);
+  StringRef GpuArch = DriverArgs.getLastArgValue(options::OPT_mcpu_EQ);
   assert(!GpuArch.empty() && "Must have an explicit GPU arch.");
   (void) GpuArch;
   assert(DeviceOffloadingKind == Action::OFK_HIP &&
@@ -294,8 +134,6 @@
   auto Kind = llvm::AMDGPU::parseArchAMDGCN(GpuArch);
   const StringRef CanonArch = llvm::AMDGPU::getArchNameAMDGCN(Kind);
 
-  CC1Args.push_back("-target-cpu");
-  CC1Args.push_back(DriverArgs.MakeArgStringRef(GpuArch));
   CC1Args.push_back("-fcuda-is-device");
 
   if (DriverArgs.hasFlag(options::OPT_fcuda_approx_transcendentals,
@@ -396,8 +234,8 @@
   }
 
   if (!BoundArch.empty()) {
-    DAL->eraseArg(options::OPT_march_EQ);
-    DAL->AddJoinedArg(nullptr, Opts.getOption(options::OPT_march_EQ), BoundArch);
+    DAL->eraseArg(options::OPT_mcpu_EQ);
+    DAL->AddJoinedArg(nullptr, Opts.getOption(options::OPT_mcpu_EQ), BoundArch);
   }
 
   return DAL;
Index: clang/test/Driver/amdgpu-toolchain.c
===================================================================
--- clang/test/Driver/amdgpu-toolchain.c
+++ clang/test/Driver/amdgpu-toolchain.c
@@ -9,3 +9,8 @@
 // AS_LINK: ld.lld{{.*}} "-shared"
 
 // DWARF_VER: "-dwarf-version=4"
+
+// RUN: %clang -### -target amdgcn-amd-amdhsa -mcpu=gfx906 -nogpulib \
+// RUN:   -flto %s 2>&1 | FileCheck -check-prefix=LTO %s
+// LTO: clang{{.*}} "-flto"
+// LTO: ld.lld{{.*}}
Index: clang/test/Driver/hip-device-compile.hip
===================================================================
--- clang/test/Driver/hip-device-compile.hip
+++ clang/test/Driver/hip-device-compile.hip
@@ -31,9 +31,10 @@
 // BC-SAME: "-emit-llvm-bc"
 // LL-SAME: "-emit-llvm"
 // ASM-NOT: "-emit-llvm"
-// CHECK-SAME: "-main-file-name" "a.cu" {{.*}} "-target-cpu" "gfx900"
+// CHECK-SAME: "-main-file-name" "a.cu"
 // CHECK-SAME: "-fcuda-is-device"
 // CHECK-SAME: {{".*lib1.bc"}}
+// CHECK-SAME: "-target-cpu" "gfx900"
 // BC-SAME: "-o" "a.bc"
 // LL-SAME: "-o" "a.ll"
 // ASM-SAME: "-o" "a.s"
Index: clang/test/Driver/hip-phases.hip
===================================================================
--- clang/test/Driver/hip-phases.hip
+++ clang/test/Driver/hip-phases.hip
@@ -30,7 +30,8 @@
 // NRD-DAG: [[P8:[0-9]+]]: linker, {[[P7]]}, image, (device-hip, [[ARCH]])
 // NRD-DAG: [[P9:[0-9]+]]: offload, "device-hip (amdgcn-amd-amdhsa:[[ARCH]])" {[[P8]]}, image
 // NRD-DAG: [[P10:[0-9]+]]: linker, {[[P9]]}, hip-fatbin, (device-hip)
-// RDC-DAG: [[P10:[0-9]+]]: linker, {[[P5]]}, image, (device-hip, [[ARCH]])
+// RDC-DAG: [[P6:[0-9]+]]: backend, {[[P5]]}, ir, (device-[[T]], [[ARCH]])
+// RDC-DAG: [[P10:[0-9]+]]: linker, {[[P6]]}, image, (device-hip, [[ARCH]])
 
 // NRD-DAG: [[P12:[0-9]+]]: offload, "host-hip (x86_64-unknown-linux-gnu)" {[[P2]]}, "device-hip (amdgcn-amd-amdhsa)" {[[P10]]}, ir
 
@@ -59,7 +60,19 @@
 //
 // RUN: %clang -x hip -target x86_64-unknown-linux-gnu -ccc-print-phases \
 // RUN: --cuda-gpu-arch=gfx803 --cuda-gpu-arch=gfx900 %s 2>&1 \
-// RUN: | FileCheck -check-prefixes=BIN2 %s
+// RUN: | FileCheck -check-prefixes=BIN2,NRD2,CL2 %s
+
+// RUN: %clang -x hip -target x86_64-unknown-linux-gnu -ccc-print-phases \
+// RUN: --cuda-gpu-arch=gfx803 --cuda-gpu-arch=gfx900 %s -c 2>&1 \
+// RUN: | FileCheck -check-prefixes=BIN2,NRD2 %s
+
+// RUN: %clang -x hip -target x86_64-unknown-linux-gnu -ccc-print-phases \
+// RUN: --cuda-gpu-arch=gfx803 --cuda-gpu-arch=gfx900 %s -fgpu-rdc 2>&1 \
+// RUN: | FileCheck -check-prefixes=BIN2,RDC2,CL2,RCL2 %s
+
+// RUN: %clang -x hip -target x86_64-unknown-linux-gnu -ccc-print-phases \
+// RUN: --cuda-gpu-arch=gfx803 --cuda-gpu-arch=gfx900 %s -fgpu-rdc -c 2>&1 \
+// RUN: | FileCheck -check-prefixes=BIN2,RDC2,RC2 %s
 
 // BIN2-DAG: [[P0:[0-9]+]]: input, "{{.*}}hip-phases.hip", [[T:hip]], (host-[[T]])
 // BIN2-DAG: [[P1:[0-9]+]]: preprocessor, {[[P0]]}, [[T]]-cpp-output, (host-[[T]])
@@ -68,25 +81,34 @@
 // BIN2-DAG: [[P3:[0-9]+]]: input, "{{.*}}hip-phases.hip", [[T]], (device-[[T]], [[ARCH1:gfx803]])
 // BIN2-DAG: [[P4:[0-9]+]]: preprocessor, {[[P3]]}, [[T]]-cpp-output, (device-[[T]], [[ARCH1]])
 // BIN2-DAG: [[P5:[0-9]+]]: compiler, {[[P4]]}, ir, (device-[[T]], [[ARCH1]])
-// BIN2-DAG: [[P6:[0-9]+]]: backend, {[[P5]]}, assembler, (device-[[T]], [[ARCH1]])
-// BIN2-DAG: [[P7:[0-9]+]]: assembler, {[[P6]]}, object, (device-[[T]], [[ARCH1]])
-// BIN2-DAG: [[P8:[0-9]+]]: linker, {[[P7]]}, image, (device-[[T]], [[ARCH1]])
-// BIN2-DAG: [[P9:[0-9]+]]: offload, "device-hip (amdgcn-amd-amdhsa:[[ARCH1]])" {[[P8]]}, image
+// NRD2-DAG: [[P6:[0-9]+]]: backend, {[[P5]]}, assembler, (device-[[T]], [[ARCH1]])
+// RDC2-DAG: [[P6:[0-9]+]]: backend, {[[P5]]}, ir, (device-[[T]], [[ARCH1]])
+// NRD2-DAG: [[P7:[0-9]+]]: assembler, {[[P6]]}, object, (device-[[T]], [[ARCH1]])
+// NRD2-DAG: [[P8:[0-9]+]]: linker, {[[P7]]}, image, (device-[[T]], [[ARCH1]])
+// RCL2-DAG: [[P8:[0-9]+]]: linker, {[[P6]]}, image, (device-[[T]], [[ARCH1]])
+// NRD2-DAG: [[P9:[0-9]+]]: offload, "device-hip (amdgcn-amd-amdhsa:[[ARCH1]])" {[[P8]]}, image
+// RC2-DAG: [[P9:[0-9]+]]: offload, "device-hip (amdgcn-amd-amdhsa:[[ARCH1]])" {[[P6]]}, ir 
 
 // BIN2-DAG: [[P10:[0-9]+]]: input, "{{.*}}hip-phases.hip", [[T]], (device-[[T]], [[ARCH2:gfx900]])
 // BIN2-DAG: [[P11:[0-9]+]]: preprocessor, {[[P10]]}, [[T]]-cpp-output, (device-[[T]], [[ARCH2]])
 // BIN2-DAG: [[P12:[0-9]+]]: compiler, {[[P11]]}, ir, (device-[[T]], [[ARCH2]])
-// BIN2-DAG: [[P13:[0-9]+]]: backend, {[[P12]]}, assembler, (device-[[T]], [[ARCH2]])
-// BIN2-DAG: [[P14:[0-9]+]]: assembler, {[[P13]]}, object, (device-[[T]], [[ARCH2]])
-// BIN2-DAG: [[P15:[0-9]+]]: linker, {[[P14]]}, image, (device-[[T]], [[ARCH2]])
-// BIN2-DAG: [[P16:[0-9]+]]: offload, "device-hip (amdgcn-amd-amdhsa:[[ARCH2]])" {[[P15]]}, image
+// NRD2-DAG: [[P13:[0-9]+]]: backend, {[[P12]]}, assembler, (device-[[T]], [[ARCH2]])
+// RDC2-DAG: [[P13:[0-9]+]]: backend, {[[P12]]}, ir, (device-[[T]], [[ARCH2]])
+// NRD2-DAG: [[P14:[0-9]+]]: assembler, {[[P13]]}, object, (device-[[T]], [[ARCH2]])
+// NRD2-DAG: [[P15:[0-9]+]]: linker, {[[P14]]}, image, (device-[[T]], [[ARCH2]])
+// RCL2-DAG: [[P15:[0-9]+]]: linker, {[[P13]]}, image, (device-[[T]], [[ARCH2]])
+// NRD2-DAG: [[P16:[0-9]+]]: offload, "device-hip (amdgcn-amd-amdhsa:[[ARCH2]])" {[[P15]]}, image
+// RC2-DAG: [[P16:[0-9]+]]: offload, "device-hip (amdgcn-amd-amdhsa:[[ARCH2]])" {[[P13]]}, ir 
 
-// BIN2-DAG: [[P17:[0-9]+]]: linker, {[[P9]], [[P16]]}, hip-fatbin, (device-hip)
+// NRD2-DAG: [[P17:[0-9]+]]: linker, {[[P9]], [[P16]]}, hip-fatbin, (device-hip)
 
-// BIN2-DAG: [[P18:[0-9]+]]: offload, "host-hip (x86_64-unknown-linux-gnu)" {[[P2]]}, "device-hip (amdgcn-amd-amdhsa)" {[[P17]]}, ir
-// BIN2-DAG: [[P19:[0-9]+]]: backend, {[[P18]]}, assembler, (host-[[T]])
+// NRD2-DAG: [[P18:[0-9]+]]: offload, "host-hip (x86_64-unknown-linux-gnu)" {[[P2]]}, "device-hip (amdgcn-amd-amdhsa)" {[[P17]]}, ir
+// NRD2-DAG: [[P19:[0-9]+]]: backend, {[[P18]]}, assembler, (host-[[T]])
+// RDC2-DAG: [[P19:[0-9]+]]: backend, {[[P2]]}, assembler, (host-[[T]])
 // BIN2-DAG: [[P20:[0-9]+]]: assembler, {[[P19]]}, object, (host-[[T]])
-// BIN2-DAG: [[P21:[0-9]+]]: linker, {[[P20]]}, image, (host-[[T]])
+// CL2-DAG: [[P21:[0-9]+]]: linker, {[[P20]]}, image, (host-[[T]])
+// RCL2-DAG: [[P22:[0-9]+]]: offload, "host-hip (x86_64-unknown-linux-gnu)" {[[P21]]}, "device-hip (amdgcn-amd-amdhsa:[[ARCH1]])" {[[P8]]}, "device-hip (amdgcn-amd-amdhsa:[[ARCH2]])" {[[P15]]}, image
+// RC2-DAG: [[P22:[0-9]+]]: clang-offload-bundler, {[[P9]], [[P16]], [[P20]]}, object, (host-[[T]])
 
 //
 // Test two gpu architecturess up to the assemble phase.
@@ -212,3 +234,27 @@
 // DASM2-DAG: [[P6:[0-9]+]]: preprocessor, {[[P5]]}, [[T]]-cpp-output, (device-[[T]], [[ARCH2]])
 // DASM2-DAG: [[P7:[0-9]+]]: compiler, {[[P6]]}, ir, (device-[[T]], [[ARCH2]])
 // DASM2-NOT: host
+
+//
+// Test linking two objects with two gpu architectures.
+//
+// RUN: touch %T/obj1.o
+// RUN: touch %T/obj2.o
+//
+// RUN: %clang -target x86_64-unknown-linux-gnu -ccc-print-phases --hip-link \
+// RUN: --cuda-gpu-arch=gfx803 --cuda-gpu-arch=gfx900 %T/obj1.o %T/obj2.o 2>&1 \
+// RUN: | FileCheck -check-prefixes=L2,NL2 %s
+//
+// RUN: %clang -target x86_64-unknown-linux-gnu -ccc-print-phases --hip-link \
+// RUN: --cuda-gpu-arch=gfx803 --cuda-gpu-arch=gfx900 %T/obj1.o %T/obj2.o \
+// RUN: -fgpu-rdc 2>&1 | FileCheck -check-prefixes=L2,RL2 %s
+//
+// L2-DAG: [[P0:[0-9]+]]: input, "{{.*}}obj1.o", object, (host-[[T:hip]])
+// RL2-DAG: [[P1:[0-9]+]]: clang-offload-unbundler, {[[P0]]}, object, (host-[[T:hip]])
+// L2-DAG: [[P2:[0-9]+]]: input, "{{.*}}obj2.o", object, (host-[[T:hip]])
+// RL2-DAG: [[P3:[0-9]+]]: clang-offload-unbundler, {[[P2]]}, object, (host-[[T:hip]])
+// NL2-DAG: [[P4:[0-9]+]]: linker, {[[P0]], [[P2]]}, image, (host-[[T:hip]])
+// RL2-DAG: [[P4:[0-9]+]]: linker, {[[P1]], [[P3]]}, image, (host-[[T:hip]])
+// RL2-DAG: [[P5:[0-9]+]]: linker, {[[P1]], [[P3]]}, image, (device-[[T:hip]], [[ARCH1:gfx803]])
+// RL2-DAG: [[P6:[0-9]+]]: linker, {[[P1]], [[P3]]}, image, (device-[[T:hip]], [[ARCH2:gfx900]])
+// RL2-DAG: [[P7:[0-9]+]]: offload, "host-[[T:hip]] (x86_64-unknown-linux-gnu)" {[[P4]]}, "device-[[T:hip]] (amdgcn-amd-amdhsa:[[ARCH1]])" {[[P5]]}, "device-[[T:hip]] (amdgcn-amd-amdhsa:[[ARCH2]])" {[[P6]]}, image
Index: clang/test/Driver/hip-rdc-device-only.hip
===================================================================
--- clang/test/Driver/hip-rdc-device-only.hip
+++ clang/test/Driver/hip-rdc-device-only.hip
@@ -51,9 +51,10 @@
 // COMMON-SAME: "-aux-triple" "x86_64-unknown-linux-gnu"
 // EMITBC-SAME: "-emit-llvm-bc"
 // EMITLL-SAME: "-emit-llvm"
-// COMMON-SAME: {{.*}} "-main-file-name" "a.cu" {{.*}} "-target-cpu" "gfx803"
+// COMMON-SAME: {{.*}} "-main-file-name" "a.cu"
 // COMMON-SAME: "-fcuda-is-device" "-fgpu-rdc" "-fcuda-allow-variadic-functions" "-fvisibility" "hidden"
 // COMMON-SAME: "-fapply-global-visibility-to-externs"
+// COMMON-SAME: "-target-cpu" "gfx803"
 // EMITBC-SAME: {{.*}} "-o" {{"a.*bc"}} "-x" "hip"
 // EMITLL-SAME: {{.*}} "-o" {{"a.*ll"}} "-x" "hip"
 // CHECK-SAME: {{.*}} {{".*a.cu"}}
@@ -62,9 +63,10 @@
 // COMMON-SAME: "-aux-triple" "x86_64-unknown-linux-gnu"
 // EMITBC-SAME: "-emit-llvm-bc"
 // EMITLL-SAME: "-emit-llvm"
-// COMMON-SAME: {{.*}} "-main-file-name" "a.cu" {{.*}} "-target-cpu" "gfx900"
+// COMMON-SAME: {{.*}} "-main-file-name" "a.cu"
 // COMMON-SAME: "-fcuda-is-device" "-fgpu-rdc" "-fcuda-allow-variadic-functions" "-fvisibility" "hidden"
 // COMMON-SAME: "-fapply-global-visibility-to-externs"
+// COMMON-SAME: "-target-cpu" "gfx900"
 // EMITBC-SAME: {{.*}} "-o" {{"a.*bc"}} "-x" "hip"
 // EMITLL-SAME: {{.*}} "-o" {{"a.*ll"}} "-x" "hip"
 // COMMON-SAME: {{.*}} {{".*a.cu"}}
@@ -73,9 +75,10 @@
 // COMMON-SAME: "-aux-triple" "x86_64-unknown-linux-gnu"
 // EMITBC-SAME: "-emit-llvm-bc"
 // EMITLL-SAME: "-emit-llvm"
-// COMMON-SAME: {{.*}} "-main-file-name" "b.hip" {{.*}} "-target-cpu" "gfx803"
+// COMMON-SAME: {{.*}} "-main-file-name" "b.hip"
 // COMMON-SAME: "-fcuda-is-device" "-fgpu-rdc" "-fcuda-allow-variadic-functions" "-fvisibility" "hidden"
 // COMMON-SAME: "-fapply-global-visibility-to-externs"
+// COMMON-SAME: "-target-cpu" "gfx803"
 // EMITBC-SAME: {{.*}} "-o" {{"b.*bc"}} "-x" "hip"
 // EMITLL-SAME: {{.*}} "-o" {{"b.*ll"}} "-x" "hip"
 // COMMON-SAME: {{.*}} {{".*b.hip"}}
@@ -84,9 +87,10 @@
 // COMMON-SAME: "-aux-triple" "x86_64-unknown-linux-gnu"
 // EMITBC-SAME: "-emit-llvm-bc"
 // EMITLL-SAME: "-emit-llvm"
-// COMMON-SAME: {{.*}} "-main-file-name" "b.hip" {{.*}} "-target-cpu" "gfx900"
+// COMMON-SAME: {{.*}} "-main-file-name" "b.hip"
 // COMMON-SAME: "-fcuda-is-device" "-fgpu-rdc" "-fcuda-allow-variadic-functions" "-fvisibility" "hidden"
 // COMMON-SAME: "-fapply-global-visibility-to-externs"
+// COMMON-SAME: "-target-cpu" "gfx900"
 // EMITBC-SAME: {{.*}} "-o" {{"b.*bc"}} "-x" "hip"
 // EMITLL-SAME: {{.*}} "-o" {{"b.*ll"}} "-x" "hip"
 // COMMON-SAME: {{.*}} {{".*b.hip"}}
Index: clang/test/Driver/hip-save-temps.hip
===================================================================
--- clang/test/Driver/hip-save-temps.hip
+++ clang/test/Driver/hip-save-temps.hip
@@ -2,6 +2,11 @@
 // REQUIRES: x86-registered-target
 // REQUIRES: amdgpu-registered-target
 
+// -fno-gpu-rdc without -o with -c
+// RUN: %clang -### -target x86_64-linux-gnu -nogpulib -save-temps \
+// RUN:   --cuda-gpu-arch=gfx900 -c %s 2>&1 | \
+// RUN:   FileCheck -check-prefixes=CHECK,NORDC %s
+
 // -fno-gpu-rdc without -o
 // RUN: %clang -### -target x86_64-linux-gnu -nogpulib -save-temps \
 // RUN:   --cuda-gpu-arch=gfx900 %s 2>&1 | \
@@ -12,37 +17,58 @@
 // RUN:   -o executable --cuda-gpu-arch=gfx900 %s 2>&1 | \
 // RUN:   FileCheck -check-prefixes=CHECK,NORDC,WOUT %s
 
+// -fgpu-rdc without -o with -c
+// RUN: %clang -### -target x86_64-linux-gnu -nogpulib -save-temps \
+// RUN:   -fgpu-rdc --cuda-gpu-arch=gfx900 -c %s 2>&1 | \
+// RUN:   FileCheck -check-prefixes=CHECK,RDC,RDCC %s
+
 // -fgpu-rdc without -o
 // RUN: %clang -### -target x86_64-linux-gnu -nogpulib -save-temps \
 // RUN:   -fgpu-rdc --cuda-gpu-arch=gfx900 %s 2>&1 | \
-// RUN:   FileCheck -check-prefixes=CHECK,RDC,RDC-NOUT,NOUT %s
+// RUN:   FileCheck -check-prefixes=CHECK,RDC,RDCL,RDC-NOUT,NOUT %s
 
 // -fgpu-rdc with -o
 // RUN: %clang -### -target x86_64-linux-gnu -nogpulib -save-temps \
 // RUN:   -o executable -fgpu-rdc --cuda-gpu-arch=gfx900 %s 2>&1 | \
-// RUN:   FileCheck -check-prefixes=CHECK,RDC,RDC-WOUT,WOUT %s
-
-// CHECK: {{.*}}clang{{.*}}"-o" "hip-save-temps-hip-amdgcn-amd-amdhsa-gfx900.cui"
-// NORDC: {{".*clang.*"}} {{.*}} "-disable-llvm-passes" {{.*}} "-o" "hip-save-temps-hip-amdgcn-amd-amdhsa-gfx900.bc"
-// NORDC: {{".*clang.*"}} {{.*}} "-S" {{.*}} "-o" "hip-save-temps-hip-amdgcn-amd-amdhsa-gfx900.s"
-// NORDC: {{".*clang.*"}} "-cc1as" {{.*}} "-filetype" "obj" {{.*}} "-o" "hip-save-temps-hip-amdgcn-amd-amdhsa-gfx900.o"
-// NORDC-NOT: {{.*}}llvm-link
-// NORDC-NOT: {{.*}}opt
-// NORDC-NOT: {{.*}}llc
-// RDC: {{.*}}llvm-link{{.*}}"-o" "hip-save-temps-hip-amdgcn-amd-amdhsa-gfx900.tmp-linked.bc"
-// RDC: {{.*}}opt{{.*}}"-o" "hip-save-temps-hip-amdgcn-amd-amdhsa-gfx900.tmp-optimized.bc"
-// RDC: {{.*}}llc{{.*}}"-filetype=asm"{{.*}}"-o" "hip-save-temps-hip-amdgcn-amd-amdhsa-gfx900.tmp.s"
-// RDC: {{.*}}llc{{.*}}"-filetype=obj"{{.*}}"-o" "hip-save-temps-hip-amdgcn-amd-amdhsa-gfx900.tmp.o"
-// NORDC: {{.*}}lld{{.*}}"-o" "hip-save-temps-hip-amdgcn-amd-amdhsa-gfx900.out"
-// RDC: {{.*}}lld{{.*}}"-o" "a.out-hip-amdgcn-amd-amdhsa-gfx900"
-// NORDC: {{.*}}clang-offload-bundler{{.*}}"-outputs=hip-save-temps.hip-hip-amdgcn-amd-amdhsa.hipfb"
-// CHECK: {{.*}}clang{{.*}}"-o" "hip-save-temps-host-x86_64-unknown-linux-gnu.cui"
-// NORDC: {{.*}}clang{{.*}}"-fcuda-include-gpubinary" {{.*}} "-o" "hip-save-temps-host-x86_64-unknown-linux-gnu.bc"
-// RDC: {{.*}}clang{{.*}}"-o" "hip-save-temps-host-x86_64-unknown-linux-gnu.bc"
-// CHECK: {{.*}}clang{{.*}}"-o" "hip-save-temps-host-x86_64-unknown-linux-gnu.s"
-// CHECK: {{.*}}clang{{.*}}"-o" "hip-save-temps{{.*}}.o"
-// RDC-NOUT: {{.*}}clang-offload-bundler{{.*}}"-outputs=a.out.hipfb"
-// RDC-WOUT: {{.*}}clang-offload-bundler{{.*}}"-outputs=executable.hipfb"
-// NOUT: {{.*}}ld{{.*}}"-o" "a.out"
-// WOUT: {{.*}}ld{{.*}}"-o" "executable"
+// RUN:   FileCheck -check-prefixes=CHECK,RDC,RDCL,RDC-WOUT,WOUT %s
+
+// CHECK: {{".*clang.*"}} "-cc1" {{.*}} "-E" {{.*}} [[CPU:"-target-cpu" "gfx900"]] {{.*}}  "-o" "hip-save-temps-hip-amdgcn-amd-amdhsa-gfx900.cui"
+// NORDC: {{".*clang.*"}} "-cc1" {{.*}} "-emit-llvm-bc" {{.*}} [[CPU]] {{.*}} "-disable-llvm-passes" {{.*}} "-o" "hip-save-temps-hip-amdgcn-amd-amdhsa-gfx900.bc"
+// RDC: {{".*clang.*"}} "-cc1" {{.*}} "-emit-llvm-bc" {{.*}} [[CPU]] {{.*}} "-disable-llvm-passes" {{.*}} "-o" "hip-save-temps-hip-amdgcn-amd-amdhsa-gfx900.tmp.bc"
+// NORDC: {{".*clang.*"}} "-cc1" {{.*}} "-S" {{.*}} [[CPU]] {{.*}} "-o" "hip-save-temps-hip-amdgcn-amd-amdhsa-gfx900.s"
+// RDC: {{".*clang.*"}} "-cc1" {{.*}} "-emit-llvm-bc" {{.*}} [[CPU]] {{.*}} "-o" "hip-save-temps-hip-amdgcn-amd-amdhsa-gfx900.bc"
+// NORDC: {{".*clang.*"}} "-cc1as" {{.*}} "-filetype" "obj" {{.*}} [[CPU]] {{.*}} "-o" "hip-save-temps-hip-amdgcn-amd-amdhsa-gfx900.o"
+// CHECK-NOT: llvm-link
+// CHECK-NOT: opt
+// CHECK-NOT: llc
+// NORDC: {{.*lld.*}}"-o" "hip-save-temps-hip-amdgcn-amd-amdhsa-gfx900.out"
+// RDCL: "{{.*lld.*}}" {{.*}} "-o" "a.out-hip-amdgcn-amd-amdhsa-gfx900"
+// NORDC: "{{.*clang-offload-bundler.*}}" {{.*}} "-outputs=hip-save-temps.hip-hip-amdgcn-amd-amdhsa.hipfb"
+// CHECK: "{{.*clang.*}}" "-cc1" {{.*}} "-E" {{.*}} "-o" "hip-save-temps-host-x86_64-unknown-linux-gnu.cui"
+// NORDC: "{{.*clang.*}}" "-cc1" {{.*}} "-emit-llvm-bc" {{.*}}  "-fcuda-include-gpubinary" {{.*}} "-o" "hip-save-temps-host-x86_64-unknown-linux-gnu.bc"
+// RDC: "{{.*clang.*}}" "-cc1" {{.*}} "-emit-llvm-bc" {{.*}} "-o" "hip-save-temps-host-x86_64-unknown-linux-gnu.bc"
+// CHECK: "{{.*clang.*}}" "-cc1" {{.*}} "-S" {{.*}} "-o" "hip-save-temps-host-x86_64-unknown-linux-gnu.s"
+// CHECK: "{{.*clang.*}}" "-cc1as" {{.*}} "-o" "hip-save-temps{{.*}}.o"
+// RDCC: "{{.*clang-offload-bundler.*}}" {{.*}} "-outputs=hip-save-temps.o"
+// RDC-NOUT: "{{.*clang-offload-bundler.*}}" {{.*}} "-outputs=a.out.hipfb"
+// RDC-WOUT: "{{.*clang-offload-bundler.*}}" {{.*}} "-outputs=executable.hipfb"
+// NOUT: "{{.*ld.*}}" {{.*}} "-o" "a.out"
+// WOUT: "{{.*ld.*}}" {{.*}} "-o" "executable"
+
+// -fgpu-rdc link
+// RUN: touch %T/obj1.o
+// RUN: touch %T/obj2.o
+// RUN: %clang -### -target x86_64-linux-gnu -nogpulib -save-temps \
+// RUN:   --hip-link -o executable -fgpu-rdc --cuda-gpu-arch=gfx900 \
+// RUN:   --offload-arch=gfx906  %T/obj1.o %T/obj2.o 2>&1 | \
+// RUN:   FileCheck -check-prefixes=LINK %s
 
+// LINK: "{{.*clang-offload-bundler.*}}" {{.*}} "-outputs=obj1-host-x86_64-unknown-linux-gnu.o,obj1-hip-amdgcn-amd-amdhsa-gfx900.o,obj1-hip-amdgcn-amd-amdhsa-gfx906.o" "-unbundle"
+// LINK: "{{.*clang-offload-bundler.*}}" {{.*}} "-outputs=obj2-host-x86_64-unknown-linux-gnu.o,obj2-hip-amdgcn-amd-amdhsa-gfx900.o,obj2-hip-amdgcn-amd-amdhsa-gfx906.o" "-unbundle"
+// LINK-NOT: llvm-link
+// LINK-NOT: opt
+// LINK-NOT: llc
+// LINK: "{{.*lld.*}}" {{.*}} "-o" "a.out-hip-amdgcn-amd-amdhsa-gfx900" "obj1-hip-amdgcn-amd-amdhsa-gfx900.o" "obj2-hip-amdgcn-amd-amdhsa-gfx900.o"
+// LINK: "{{.*lld.*}}" {{.*}} "-o" "a.out-hip-amdgcn-amd-amdhsa-gfx906" "obj1-hip-amdgcn-amd-amdhsa-gfx906.o" "obj2-hip-amdgcn-amd-amdhsa-gfx906.o"
+// LINK: "{{.*clang-offload-bundler.*}}" {{.*}} "-outputs=executable.hipfb"
+// LINK: "{{.*ld.*}}" {{.*}} "-o" "executable" {{.*}} "-T" "executable.lk"
Index: clang/test/Driver/hip-toolchain-no-rdc.hip
===================================================================
--- clang/test/Driver/hip-toolchain-no-rdc.hip
+++ clang/test/Driver/hip-toolchain-no-rdc.hip
@@ -7,11 +7,29 @@
 // RUN:   --hip-device-lib=lib1.bc --hip-device-lib=lib2.bc \
 // RUN:   --hip-device-lib-path=%S/Inputs/hip_multiple_inputs/lib1 \
 // RUN:   --hip-device-lib-path=%S/Inputs/hip_multiple_inputs/lib2 \
-// RUN:   -fuse-ld=lld \
+// RUN:   -fuse-ld=lld -nogpuinc \
+// RUN:   %S/Inputs/hip_multiple_inputs/a.cu \
+// RUN:   %S/Inputs/hip_multiple_inputs/b.hip \
+// RUN: 2>&1 | FileCheck -check-prefixes=CHECK,LINK %s
+
+// RUN: %clang -### -target x86_64-linux-gnu -fno-gpu-rdc \
+// RUN:   -x hip --cuda-gpu-arch=gfx803 --cuda-gpu-arch=gfx900 \
+// RUN:   --hip-device-lib=lib1.bc --hip-device-lib=lib2.bc \
+// RUN:   --hip-device-lib-path=%S/Inputs/hip_multiple_inputs/lib1 \
+// RUN:   --hip-device-lib-path=%S/Inputs/hip_multiple_inputs/lib2 \
+// RUN:   -fuse-ld=lld -nogpuinc -c \
 // RUN:   %S/Inputs/hip_multiple_inputs/a.cu \
 // RUN:   %S/Inputs/hip_multiple_inputs/b.hip \
 // RUN: 2>&1 | FileCheck -check-prefixes=CHECK %s
 
+// RUN: touch %T/a.o
+// RUN: touch %T/b.o
+// RUN: %clang -### -target x86_64-linux-gnu \
+// RUN:   --hip-link --cuda-gpu-arch=gfx803 --cuda-gpu-arch=gfx900 \
+// RUN:   -fuse-ld=lld -nogpuinc \
+// RUN:   %T/a.o %T/b.o \
+// RUN: 2>&1 | FileCheck -check-prefixes=LKONLY %s
+
 //
 // Compile device code in a.cu to code object for gfx803.
 //
@@ -19,10 +37,11 @@
 // CHECK: [[CLANG:".*clang.*"]] "-cc1" "-triple" "amdgcn-amd-amdhsa"
 // CHECK-SAME: "-aux-triple" "x86_64-unknown-linux-gnu"
 // CHECK-SAME: "-emit-obj"
-// CHECK-SAME: {{.*}} "-main-file-name" "a.cu" {{.*}} "-target-cpu" "gfx803"
+// CHECK-SAME: {{.*}} "-main-file-name" "a.cu"
 // CHECK-SAME: "-fcuda-is-device" "-fcuda-allow-variadic-functions" "-fvisibility" "hidden"
 // CHECK-SAME: "-fapply-global-visibility-to-externs"
 // CHECK-SAME: "{{.*}}lib1.bc" "{{.*}}lib2.bc"
+// CHECK-SAME: "-target-cpu" "gfx803"
 // CHECK-SAME: {{.*}} "-o" [[OBJ_DEV_A_803:".*o"]] "-x" "hip"
 // CHECK-SAME: {{.*}} [[A_SRC:".*a.cu"]]
 
@@ -40,10 +59,11 @@
 // CHECK: [[CLANG:".*clang.*"]] "-cc1" "-triple" "amdgcn-amd-amdhsa"
 // CHECK-SAME: "-aux-triple" "x86_64-unknown-linux-gnu"
 // CHECK-SAME: "-emit-obj"
-// CHECK-SAME: {{.*}} "-main-file-name" "a.cu" {{.*}} "-target-cpu" "gfx900"
+// CHECK-SAME: {{.*}} "-main-file-name" "a.cu"
 // CHECK-SAME: "-fcuda-is-device" "-fcuda-allow-variadic-functions" "-fvisibility" "hidden"
 // CHECK-SAME: "-fapply-global-visibility-to-externs"
 // CHECK-SAME: "{{.*}}lib1.bc" "{{.*}}lib2.bc"
+// CHECK-SAME: "-target-cpu" "gfx900"
 // CHECK-SAME: {{.*}} "-o" [[OBJ_DEV_A_900:".*o"]] "-x" "hip"
 // CHECK-SAME: {{.*}} [[A_SRC]]
 
@@ -77,10 +97,11 @@
 // CHECK: [[CLANG:".*clang.*"]] "-cc1" "-triple" "amdgcn-amd-amdhsa"
 // CHECK-SAME: "-aux-triple" "x86_64-unknown-linux-gnu"
 // CHECK-SAME: "-emit-obj"
-// CHECK-SAME: {{.*}} "-main-file-name" "b.hip" {{.*}} "-target-cpu" "gfx803"
+// CHECK-SAME: {{.*}} "-main-file-name" "b.hip"
 // CHECK-SAME: "-fcuda-is-device" "-fcuda-allow-variadic-functions" "-fvisibility" "hidden"
 // CHECK-SAME: "-fapply-global-visibility-to-externs"
 // CHECK-SAME: "{{.*}}lib1.bc" "{{.*}}lib2.bc"
+// CHECK-SAME: "-target-cpu" "gfx803"
 // CHECK-SAME: {{.*}} "-o" [[OBJ_DEV_B_803:".*o"]] "-x" "hip"
 // CHECK-SAME: {{.*}} [[B_SRC:".*b.hip"]]
 
@@ -98,10 +119,11 @@
 // CHECK: [[CLANG:".*clang.*"]] "-cc1" "-triple" "amdgcn-amd-amdhsa"
 // CHECK-SAME: "-aux-triple" "x86_64-unknown-linux-gnu"
 // CHECK-SAME: "-emit-obj"
-// CHECK-SAME: {{.*}} "-main-file-name" "b.hip" {{.*}} "-target-cpu" "gfx900"
+// CHECK-SAME: {{.*}} "-main-file-name" "b.hip"
 // CHECK-SAME: "-fcuda-is-device" "-fcuda-allow-variadic-functions" "-fvisibility" "hidden"
 // CHECK-SAME: "-fapply-global-visibility-to-externs"
 // CHECK-SAME: "{{.*}}lib1.bc" "{{.*}}lib2.bc"
+// CHECK-SAME: "-target-cpu" "gfx900"
 // CHECK-SAME: {{.*}} "-o" [[OBJ_DEV_B_900:".*o"]] "-x" "hip"
 // CHECK-SAME: {{.*}} [[B_SRC]]
 
@@ -132,5 +154,14 @@
 // Link host objects.
 //
 
-// CHECK: [[LD:".*ld.*"]] {{.*}} [[A_OBJ_HOST]] [[B_OBJ_HOST]]
-// CHECK-NOT: "-T" "{{.*}}.lk"
+// LINK-NOT: llvm-link
+// LINK-NOT: opt
+// LINK-NOT: llc
+// LINK: [[LD:".*ld.*"]] {{.*}} [[A_OBJ_HOST]] [[B_OBJ_HOST]]
+// LINK-NOT: "-T" "{{.*}}.lk"
+
+// LKONLY-NOT: llvm-link
+// LKONLY-NOT: opt
+// LKONLY-NOT: llc
+// LKONLY: [[LD:".*ld.*"]] {{.*}} "{{.*/a.o}}" "{{.*/b.o}}"
+// LKONLY-NOT: "-T" "{{.*}}.lk"
Index: clang/test/Driver/hip-toolchain-rdc-separate.hip
===================================================================
--- /dev/null
+++ clang/test/Driver/hip-toolchain-rdc-separate.hip
@@ -0,0 +1,112 @@
+// REQUIRES: clang-driver
+// REQUIRES: x86-registered-target
+// REQUIRES: amdgpu-registered-target
+
+// RUN: %clang -c -### -target x86_64-linux-gnu \
+// RUN:   -x hip --cuda-gpu-arch=gfx803 --cuda-gpu-arch=gfx900 \
+// RUN:   --hip-device-lib=lib1.bc --hip-device-lib=lib2.bc \
+// RUN:   --hip-device-lib-path=%S/Inputs/hip_multiple_inputs/lib1 \
+// RUN:   --hip-device-lib-path=%S/Inputs/hip_multiple_inputs/lib2 \
+// RUN:   -fuse-ld=lld -fgpu-rdc -nogpuinc \
+// RUN:   %S/Inputs/hip_multiple_inputs/a.cu \
+// RUN:   %S/Inputs/hip_multiple_inputs/b.hip \
+// RUN: 2>&1 | FileCheck %s
+
+// CHECK: [[CLANG:".*clang.*"]] "-cc1" "-triple" "amdgcn-amd-amdhsa"
+// CHECK-SAME: "-aux-triple" "x86_64-unknown-linux-gnu"
+// CHECK-SAME: "-emit-llvm-bc"
+// CHECK-SAME: {{.*}} "-main-file-name" "a.cu"
+// CHECK-SAME: "-fcuda-is-device" "-fgpu-rdc" "-fcuda-allow-variadic-functions" "-fvisibility" "hidden"
+// CHECK-SAME: "-fapply-global-visibility-to-externs"
+// CHECK-SAME: "{{.*}}lib1.bc" "{{.*}}lib2.bc"
+// CHECK-SAME: "-target-cpu" "gfx803"
+// CHECK-SAME: {{.*}} "-o" "[[A_BC1:.*bc]]" "-x" "hip"
+// CHECK-SAME: {{.*}} [[A_SRC:".*a.cu"]]
+
+// CHECK: [[CLANG]] "-cc1" "-triple" "amdgcn-amd-amdhsa"
+// CHECK-SAME: "-aux-triple" "x86_64-unknown-linux-gnu"
+// CHECK-SAME: "-emit-llvm-bc"
+// CHECK-SAME: {{.*}} "-main-file-name" "a.cu"
+// CHECK-SAME: "-fcuda-is-device" "-fgpu-rdc"
+// CHECK-SAME: "{{.*}}lib1.bc" "{{.*}}lib2.bc"
+// CHECK-SAME: "-target-cpu" "gfx900"
+// CHECK-SAME: {{.*}} "-o" "[[A_BC2:.*bc]]" "-x" "hip"
+// CHECK-SAME: {{.*}} [[A_SRC]]
+
+// CHECK: [[CLANG]] "-cc1" "-triple" "x86_64-unknown-linux-gnu"
+// CHECK-SAME: "-aux-triple" "amdgcn-amd-amdhsa"
+// CHECK-SAME: "-emit-obj"
+// CHECK-SAME: {{.*}} "-main-file-name" "a.cu"
+// CHECK-SAME: {{.*}} "-o" "[[A_OBJ_HOST:.*o]]" "-x" "hip"
+// CHECK-SAME: {{.*}} [[A_SRC]]
+
+// CHECK: [[BUNDLER:".*clang-offload-bundler"]] "-type=o"
+// CHECK-SAME: "-targets=hip-amdgcn-amd-amdhsa-gfx803,hip-amdgcn-amd-amdhsa-gfx900,host-x86_64-unknown-linux-gnu"
+// CHECK-SAME: "-outputs=[[A_O:.*a.o]]" "-inputs=[[A_BC1]],[[A_BC2]],[[A_OBJ_HOST]]"
+
+// CHECK: [[CLANG]] "-cc1" "-triple" "amdgcn-amd-amdhsa"
+// CHECK-SAME: "-aux-triple" "x86_64-unknown-linux-gnu"
+// CHECK-SAME: "-emit-llvm-bc"
+// CHECK-SAME: {{.*}} "-main-file-name" "b.hip"
+// CHECK-SAME: "-fcuda-is-device" "-fgpu-rdc" "-fcuda-allow-variadic-functions" "-fvisibility" "hidden"
+// CHECK-SAME: "-fapply-global-visibility-to-externs"
+// CHECK-SAME: "{{.*}}lib1.bc" "{{.*}}lib2.bc"
+// CHECK-SAME: "-target-cpu" "gfx803"
+// CHECK-SAME: {{.*}} "-o" "[[B_BC1:.*bc]]" "-x" "hip"
+// CHECK-SAME: {{.*}} [[B_SRC:".*b.hip"]]
+
+// CHECK: [[CLANG]] "-cc1" "-triple" "amdgcn-amd-amdhsa"
+// CHECK-SAME: "-aux-triple" "x86_64-unknown-linux-gnu"
+// CHECK-SAME: "-emit-llvm-bc"
+// CHECK-SAME: {{.*}} "-main-file-name" "b.hip"
+// CHECK-SAME: "-fcuda-is-device" "-fgpu-rdc"
+// CHECK-SAME: "{{.*}}lib1.bc" "{{.*}}lib2.bc"
+// CHECK-SAME: "-target-cpu" "gfx900"
+// CHECK-SAME: {{.*}} "-o" "[[B_BC2:.*bc]]" "-x" "hip"
+// CHECK-SAME: {{.*}} [[B_SRC]]
+
+// CHECK: [[CLANG]] "-cc1" "-triple" "x86_64-unknown-linux-gnu"
+// CHECK-SAME: "-aux-triple" "amdgcn-amd-amdhsa"
+// CHECK-SAME: "-emit-obj"
+// CHECK-SAME: {{.*}} "-main-file-name" "b.hip"
+// CHECK-SAME: {{.*}} "-o" "[[B_OBJ_HOST:.*o]]" "-x" "hip"
+// CHECK-SAME: {{.*}} [[B_SRC]]
+
+// CHECK: [[BUNDLER:".*clang-offload-bundler"]] "-type=o"
+// CHECK-SAME: "-targets=hip-amdgcn-amd-amdhsa-gfx803,hip-amdgcn-amd-amdhsa-gfx900,host-x86_64-unknown-linux-gnu"
+// CHECK-SAME: "-outputs=[[B_O:.*b.o]]" "-inputs=[[B_BC1]],[[B_BC2]],[[B_OBJ_HOST]]"
+
+// RUN: touch %T/a.o
+// RUN: touch %T/b.o
+// RUN: %clang --hip-link -### -target x86_64-linux-gnu \
+// RUN:   --cuda-gpu-arch=gfx803 --cuda-gpu-arch=gfx900 \
+// RUN:   -fuse-ld=lld -fgpu-rdc -nogpuinc \
+// RUN:   %T/a.o %T/b.o \
+// RUN: 2>&1 | FileCheck -check-prefix=LINK %s
+
+// LINK: [[BUNDLER:".*clang-offload-bundler"]] "-type=o"
+// LINK-SAME: "-targets=host-x86_64-unknown-linux-gnu,hip-amdgcn-amd-amdhsa-gfx803,hip-amdgcn-amd-amdhsa-gfx900"
+// LINK-SAME: "-inputs=[[A_O:.*a.o]]" "-outputs=[[A_OBJ_HOST:.*o]],[[A_BC1:.*o]],[[A_BC2:.*o]]"
+// LINK: "-unbundle"
+
+// LINK: [[BUNDLER:".*clang-offload-bundler"]] "-type=o"
+// LINK-SAME: "-targets=host-x86_64-unknown-linux-gnu,hip-amdgcn-amd-amdhsa-gfx803,hip-amdgcn-amd-amdhsa-gfx900"
+// LINK-SAME: "-inputs=[[B_O:.*b.o]]" "-outputs=[[B_OBJ_HOST:.*o]],[[B_BC1:.*o]],[[B_BC2:.*o]]"
+// LINK: "-unbundle"
+
+// LINK-NOT: "*.llvm-link"
+// LINK-NOT: ".*opt"
+// LINK-NOT: ".*llc"
+// LINK: {{".*lld.*"}} {{.*}} "-o" "[[IMG_DEV1:.*.out]]" "[[A_BC1]]" "[[B_BC1]]"
+
+// LINK-NOT: "*.llvm-link"
+// LINK-NOT: ".*opt"
+// LINK-NOT: ".*llc"
+// LINK: {{".*lld.*"}} {{.*}} "-o" "[[IMG_DEV2:.*.out]]" "[[A_BC2]]" "[[B_BC2]]"
+
+// LINK: [[BUNDLER:".*clang-offload-bundler"]] "-type=o"
+// LINK-SAME: "-targets={{.*}},hip-amdgcn-amd-amdhsa-gfx803,hip-amdgcn-amd-amdhsa-gfx900"
+// LINK-SAME: "-inputs={{.*}},[[IMG_DEV1]],[[IMG_DEV2]]" "-outputs=[[BUNDLE:.*hipfb]]"
+
+// LINK: [[LD:".*ld.*"]] {{.*}} "[[A_OBJ_HOST]]" "[[B_OBJ_HOST]]"
+// LINK-SAME: {{.*}} "-T" "{{.*}}.lk"
Index: clang/test/Driver/hip-toolchain-rdc.hip
===================================================================
--- clang/test/Driver/hip-toolchain-rdc.hip
+++ clang/test/Driver/hip-toolchain-rdc.hip
@@ -7,7 +7,7 @@
 // RUN:   --hip-device-lib=lib1.bc --hip-device-lib=lib2.bc \
 // RUN:   --hip-device-lib-path=%S/Inputs/hip_multiple_inputs/lib1 \
 // RUN:   --hip-device-lib-path=%S/Inputs/hip_multiple_inputs/lib2 \
-// RUN:   -fuse-ld=lld -fgpu-rdc \
+// RUN:   -fuse-ld=lld -fgpu-rdc -nogpuinc \
 // RUN:   %S/Inputs/hip_multiple_inputs/a.cu \
 // RUN:   %S/Inputs/hip_multiple_inputs/b.hip \
 // RUN: 2>&1 | FileCheck %s
@@ -15,70 +15,54 @@
 // CHECK: [[CLANG:".*clang.*"]] "-cc1" "-triple" "amdgcn-amd-amdhsa"
 // CHECK-SAME: "-aux-triple" "x86_64-unknown-linux-gnu"
 // CHECK-SAME: "-emit-llvm-bc"
-// CHECK-SAME: {{.*}} "-main-file-name" "a.cu" {{.*}} "-target-cpu" "gfx803"
+// CHECK-SAME: {{.*}} "-main-file-name" "a.cu"
 // CHECK-SAME: "-fcuda-is-device" "-fgpu-rdc" "-fcuda-allow-variadic-functions" "-fvisibility" "hidden"
 // CHECK-SAME: "-fapply-global-visibility-to-externs"
 // CHECK-SAME: "{{.*}}lib1.bc" "{{.*}}lib2.bc"
-// CHECK-SAME: {{.*}} "-o" [[A_BC:".*bc"]] "-x" "hip"
+// CHECK-SAME: "-target-cpu" "gfx803"
+// CHECK-SAME: {{.*}} "-o" [[A_BC1:".*bc"]] "-x" "hip"
 // CHECK-SAME: {{.*}} [[A_SRC:".*a.cu"]]
 
 // CHECK: [[CLANG]] "-cc1" "-triple" "amdgcn-amd-amdhsa"
 // CHECK-SAME: "-aux-triple" "x86_64-unknown-linux-gnu"
 // CHECK-SAME: "-emit-llvm-bc"
-// CHECK-SAME: {{.*}} "-main-file-name" "b.hip" {{.*}} "-target-cpu" "gfx803"
+// CHECK-SAME: {{.*}} "-main-file-name" "b.hip"
 // CHECK-SAME: "-fcuda-is-device" "-fgpu-rdc" "-fcuda-allow-variadic-functions" "-fvisibility" "hidden"
 // CHECK-SAME: "-fapply-global-visibility-to-externs"
 // CHECK-SAME: "{{.*}}lib1.bc" "{{.*}}lib2.bc"
-// CHECK-SAME: {{.*}} "-o" [[B_BC:".*bc"]] "-x" "hip"
+// CHECK-SAME: "-target-cpu" "gfx803"
+// CHECK-SAME: {{.*}} "-o" [[B_BC1:".*bc"]] "-x" "hip"
 // CHECK-SAME: {{.*}} [[B_SRC:".*b.hip"]]
 
-// CHECK: [[LLVM_LINK:"*.llvm-link"]] [[A_BC]] [[B_BC]]
-// CHECK-SAME: "-o" [[LINKED_BC_DEV1:".*-gfx803-linked-.*bc"]]
-
-// CHECK: [[OPT:".*opt"]] [[LINKED_BC_DEV1]] "-mtriple=amdgcn-amd-amdhsa"
-// CHECK-SAME: "-mcpu=gfx803"
-// CHECK-SAME: "-o" [[OPT_BC_DEV1:".*-gfx803-optimized.*bc"]]
-
-// CHECK: [[LLC: ".*llc"]] [[OPT_BC_DEV1]] "-mtriple=amdgcn-amd-amdhsa"
-// CHECK-SAME: "-mcpu=gfx803"
-// CHECK-SAME: "-filetype=obj"
-// CHECK-SAME: "-o" [[OBJ_DEV1:".*-gfx803-.*o"]]
-
-// CHECK: [[LLD: ".*lld.*"]] "-flavor" "gnu" "--no-undefined" "-shared"
-// CHECK-SAME: "-o" "[[IMG_DEV1:.*out]]" [[OBJ_DEV1]]
+// CHECK-NOT: "*.llvm-link"
+// CHECK-NOT: ".*opt"
+// CHECK-NOT: ".*llc"
+// CHECK: {{".*lld.*"}} {{.*}} "-o" "[[IMG_DEV1:.*.out]]" [[A_BC1]] [[B_BC1]]
 
 // CHECK: [[CLANG]] "-cc1" "-triple" "amdgcn-amd-amdhsa"
 // CHECK-SAME: "-aux-triple" "x86_64-unknown-linux-gnu"
 // CHECK-SAME: "-emit-llvm-bc"
-// CHECK-SAME: {{.*}} "-main-file-name" "a.cu" {{.*}} "-target-cpu" "gfx900"
+// CHECK-SAME: {{.*}} "-main-file-name" "a.cu"
 // CHECK-SAME: "-fcuda-is-device" "-fgpu-rdc"
 // CHECK-SAME: "{{.*}}lib1.bc" "{{.*}}lib2.bc"
-// CHECK-SAME: {{.*}} "-o" [[A_BC:".*bc"]] "-x" "hip"
+// CHECK-SAME: "-target-cpu" "gfx900"
+// CHECK-SAME: {{.*}} "-o" [[A_BC2:".*bc"]] "-x" "hip"
 // CHECK-SAME: {{.*}} [[A_SRC]]
 
 // CHECK: [[CLANG]] "-cc1" "-triple" "amdgcn-amd-amdhsa"
 // CHECK-SAME: "-aux-triple" "x86_64-unknown-linux-gnu"
 // CHECK-SAME: "-emit-llvm-bc"
-// CHECK-SAME: {{.*}} "-main-file-name" "b.hip" {{.*}} "-target-cpu" "gfx900"
+// CHECK-SAME: {{.*}} "-main-file-name" "b.hip"
 // CHECK-SAME: "-fcuda-is-device" "-fgpu-rdc"
 // CHECK-SAME: "{{.*}}lib1.bc" "{{.*}}lib2.bc"
-// CHECK-SAME: {{.*}} "-o" [[B_BC:".*bc"]] "-x" "hip"
+// CHECK-SAME: "-target-cpu" "gfx900"
+// CHECK-SAME: {{.*}} "-o" [[B_BC2:".*bc"]] "-x" "hip"
 // CHECK-SAME: {{.*}} [[B_SRC]]
 
-// CHECK: [[LLVM_LINK]] [[A_BC]] [[B_BC]]
-// CHECK-SAME: "-o" [[LINKED_BC_DEV2:".*-gfx900-linked-.*bc"]]
-
-// CHECK: [[OPT]] [[LINKED_BC_DEV2]] "-mtriple=amdgcn-amd-amdhsa"
-// CHECK-SAME: "-mcpu=gfx900"
-// CHECK-SAME: "-o" [[OPT_BC_DEV2:".*-gfx900-optimized.*bc"]]
-
-// CHECK: [[LLC]] [[OPT_BC_DEV2]] "-mtriple=amdgcn-amd-amdhsa"
-// CHECk-SAME: "-mcpu=gfx900"
-// CHECK-SAME: "-filetype=obj"
-// CHECK-SAME: "-o" [[OBJ_DEV2:".*-gfx900-.*o"]]
-
-// CHECK: [[LLD]] "-flavor" "gnu" "--no-undefined" "-shared"
-// CHECK-SAME: "-o" "[[IMG_DEV2:.*out]]" [[OBJ_DEV2]]
+// CHECK-NOT: "*.llvm-link"
+// CHECK-NOT: ".*opt"
+// CHECK-NOT: ".*llc"
+// CHECK: {{".*lld.*"}} {{.*}} "-o" "[[IMG_DEV2:.*.out]]" [[A_BC2]] [[B_BC2]]
 
 // CHECK: [[CLANG]] "-cc1" "-triple" "x86_64-unknown-linux-gnu"
 // CHECK-SAME: "-aux-triple" "amdgcn-amd-amdhsa"