Add basic -mtune support (#98517)

AlexisPerry · web-flow · commit f1d3fe7aae78 · 2024-07-16T16:48:24.000+01:00
Initial implementation for the -mtune flag in Flang. This PR is a clean version of PR #96688, which is a re-land of PR #95043
diff --git a/clang/include/clang/Driver/Options.td b/clang/include/clang/Driver/Options.td
@@ -5456,6 +5456,7 @@ def module_file_info : Flag<["-"], "module-file-info">, Flags<[]>,
   HelpText<"Provide information about a particular module file">;
 def mthumb : Flag<["-"], "mthumb">, Group<m_Group>;
 def mtune_EQ : Joined<["-"], "mtune=">, Group<m_Group>,
+  Visibility<[ClangOption, FlangOption]>,
   HelpText<"Only supported on AArch64, PowerPC, RISC-V, SPARC, SystemZ, and X86">;
 def multi__module : Flag<["-"], "multi_module">;
 def multiply__defined__unused : Separate<["-"], "multiply_defined_unused">;
@@ -6783,9 +6784,6 @@ def emit_hlfir : Flag<["-"], "emit-hlfir">, Group<Action_Group>,
 
 let Visibility = [CC1Option, CC1AsOption] in {
 
-def tune_cpu : Separate<["-"], "tune-cpu">,
-  HelpText<"Tune for a specific cpu type">,
-  MarshallingInfoString<TargetOpts<"TuneCPU">>;
 def target_abi : Separate<["-"], "target-abi">,
   HelpText<"Target a particular ABI type">,
   MarshallingInfoString<TargetOpts<"ABI">>;
@@ -6812,6 +6810,9 @@ def darwin_target_variant_triple : Separate<["-"], "darwin-target-variant-triple
 
 let Visibility = [CC1Option, CC1AsOption, FC1Option] in {
 
+def tune_cpu : Separate<["-"], "tune-cpu">,
+  HelpText<"Tune for a specific cpu type">,
+  MarshallingInfoString<TargetOpts<"TuneCPU">>;
 def target_cpu : Separate<["-"], "target-cpu">,
   HelpText<"Target a specific cpu type">,
   MarshallingInfoString<TargetOpts<"CPU">>;
diff --git a/clang/lib/Driver/ToolChains/Flang.cpp b/clang/lib/Driver/ToolChains/Flang.cpp
@@ -15,6 +15,7 @@
 #include "llvm/Frontend/Debug/Options.h"
 #include "llvm/Support/FileSystem.h"
 #include "llvm/Support/Path.h"
+#include "llvm/TargetParser/Host.h"
 #include "llvm/TargetParser/RISCVISAInfo.h"
 #include "llvm/TargetParser/RISCVTargetParser.h"
 
@@ -419,6 +420,13 @@ void Flang::addTargetOptions(const ArgList &Args,
   }
 
   // TODO: Add target specific flags, ABI, mtune option etc.
+  if (const Arg *A = Args.getLastArg(options::OPT_mtune_EQ)) {
+    CmdArgs.push_back("-tune-cpu");
+    if (A->getValue() == StringRef{"native"})
+      CmdArgs.push_back(Args.MakeArgString(llvm::sys::getHostCPUName()));
+    else
+      CmdArgs.push_back(A->getValue());
+  }
 }
 
 void Flang::addOffloadOptions(Compilation &C, const InputInfoList &Inputs,
@@ -815,7 +823,7 @@ void Flang::ConstructJob(Compilation &C, const JobAction &JA,
   case CodeGenOptions::FramePointerKind::None:
     FPKeepKindStr = "-mframe-pointer=none";
     break;
-   case CodeGenOptions::FramePointerKind::Reserved:
+  case CodeGenOptions::FramePointerKind::Reserved:
     FPKeepKindStr = "-mframe-pointer=reserved";
     break;
   case CodeGenOptions::FramePointerKind::NonLeaf:
diff --git a/flang/include/flang/Frontend/TargetOptions.h b/flang/include/flang/Frontend/TargetOptions.h
@@ -32,6 +32,9 @@ class TargetOptions {
   /// If given, the name of the target CPU to generate code for.
   std::string cpu;
 
+  /// If given, the name of the target CPU to tune code for.
+  std::string cpuToTuneFor;
+
   /// The list of target specific features to enable or disable, as written on
   /// the command line.
   std::vector<std::string> featuresAsWritten;
diff --git a/flang/include/flang/Lower/Bridge.h b/flang/include/flang/Lower/Bridge.h
@@ -65,11 +65,11 @@ class LoweringBridge {
          const Fortran::lower::LoweringOptions &loweringOptions,
          const std::vector<Fortran::lower::EnvironmentDefault> &envDefaults,
          const Fortran::common::LanguageFeatureControl &languageFeatures,
-         const llvm::TargetMachine &targetMachine) {
+         const llvm::TargetMachine &targetMachine, llvm::StringRef tuneCPU) {
     return LoweringBridge(ctx, semanticsContext, defaultKinds, intrinsics,
                           targetCharacteristics, allCooked, triple, kindMap,
                           loweringOptions, envDefaults, languageFeatures,
-                          targetMachine);
+                          targetMachine, tuneCPU);
   }
 
   //===--------------------------------------------------------------------===//
@@ -148,7 +148,7 @@ class LoweringBridge {
       const Fortran::lower::LoweringOptions &loweringOptions,
       const std::vector<Fortran::lower::EnvironmentDefault> &envDefaults,
       const Fortran::common::LanguageFeatureControl &languageFeatures,
-      const llvm::TargetMachine &targetMachine);
+      const llvm::TargetMachine &targetMachine, const llvm::StringRef tuneCPU);
   LoweringBridge() = delete;
   LoweringBridge(const LoweringBridge &) = delete;
 
diff --git a/flang/include/flang/Optimizer/CodeGen/CGPasses.td b/flang/include/flang/Optimizer/CodeGen/CGPasses.td
@@ -31,6 +31,8 @@ def FIRToLLVMLowering : Pass<"fir-to-llvm-ir", "mlir::ModuleOp"> {
            "Override module's data layout.">,
     Option<"forcedTargetCPU", "target-cpu", "std::string", /*default=*/"",
            "Override module's target CPU.">,
+    Option<"forcedTuneCPU", "tune-cpu", "std::string", /*default=*/"",
+           "Override module's tune CPU.">,
     Option<"forcedTargetFeatures", "target-features", "std::string",
            /*default=*/"", "Override module's target features.">,
     Option<"applyTBAA", "apply-tbaa", "bool", /*default=*/"false",
@@ -68,6 +70,8 @@ def TargetRewritePass : Pass<"target-rewrite", "mlir::ModuleOp"> {
            "Override module's target triple.">,
     Option<"forcedTargetCPU", "target-cpu", "std::string", /*default=*/"",
            "Override module's target CPU.">,
+    Option<"forcedTuneCPU", "tune-cpu", "std::string", /*default=*/"",
+           "Override module's tune CPU.">,
     Option<"forcedTargetFeatures", "target-features", "std::string",
            /*default=*/"", "Override module's target features.">,
     Option<"noCharacterConversion", "no-character-conversion",
diff --git a/flang/include/flang/Optimizer/CodeGen/Target.h b/flang/include/flang/Optimizer/CodeGen/Target.h
@@ -76,14 +76,29 @@ class CodeGenSpecifics {
       llvm::StringRef targetCPU, mlir::LLVM::TargetFeaturesAttr targetFeatures,
       const mlir::DataLayout &dl);
 
+  static std::unique_ptr<CodeGenSpecifics>
+  get(mlir::MLIRContext *ctx, llvm::Triple &&trp, KindMapping &&kindMap,
+      llvm::StringRef targetCPU, mlir::LLVM::TargetFeaturesAttr targetFeatures,
+      const mlir::DataLayout &dl, llvm::StringRef tuneCPU);
+
   static TypeAndAttr getTypeAndAttr(mlir::Type t) { return TypeAndAttr{t, {}}; }
 
   CodeGenSpecifics(mlir::MLIRContext *ctx, llvm::Triple &&trp,
                    KindMapping &&kindMap, llvm::StringRef targetCPU,
                    mlir::LLVM::TargetFeaturesAttr targetFeatures,
                    const mlir::DataLayout &dl)
       : context{*ctx}, triple{std::move(trp)}, kindMap{std::move(kindMap)},
-        targetCPU{targetCPU}, targetFeatures{targetFeatures}, dataLayout{&dl} {}
+        targetCPU{targetCPU}, targetFeatures{targetFeatures}, dataLayout{&dl},
+        tuneCPU{""} {}
+
+  CodeGenSpecifics(mlir::MLIRContext *ctx, llvm::Triple &&trp,
+                   KindMapping &&kindMap, llvm::StringRef targetCPU,
+                   mlir::LLVM::TargetFeaturesAttr targetFeatures,
+                   const mlir::DataLayout &dl, llvm::StringRef tuneCPU)
+      : context{*ctx}, triple{std::move(trp)}, kindMap{std::move(kindMap)},
+        targetCPU{targetCPU}, targetFeatures{targetFeatures}, dataLayout{&dl},
+        tuneCPU{tuneCPU} {}
+
   CodeGenSpecifics() = delete;
   virtual ~CodeGenSpecifics() {}
 
@@ -165,6 +180,7 @@ class CodeGenSpecifics {
   virtual unsigned char getCIntTypeWidth() const = 0;
 
   llvm::StringRef getTargetCPU() const { return targetCPU; }
+  llvm::StringRef getTuneCPU() const { return tuneCPU; }
 
   mlir::LLVM::TargetFeaturesAttr getTargetFeatures() const {
     return targetFeatures;
@@ -182,6 +198,7 @@ class CodeGenSpecifics {
   llvm::StringRef targetCPU;
   mlir::LLVM::TargetFeaturesAttr targetFeatures;
   const mlir::DataLayout *dataLayout = nullptr;
+  llvm::StringRef tuneCPU;
 };
 
 } // namespace fir
diff --git a/flang/include/flang/Optimizer/Dialect/Support/FIRContext.h b/flang/include/flang/Optimizer/Dialect/Support/FIRContext.h
@@ -58,6 +58,13 @@ void setTargetCPU(mlir::ModuleOp mod, llvm::StringRef cpu);
 /// Get the target CPU string from the Module or return a null reference.
 llvm::StringRef getTargetCPU(mlir::ModuleOp mod);
 
+/// Set the tune CPU for the module. `cpu` must not be deallocated while
+/// module `mod` is still live.
+void setTuneCPU(mlir::ModuleOp mod, llvm::StringRef cpu);
+
+/// Get the tune CPU string from the Module or return a null reference.
+llvm::StringRef getTuneCPU(mlir::ModuleOp mod);
+
 /// Set the target features for the module.
 void setTargetFeatures(mlir::ModuleOp mod, llvm::StringRef features);
 
diff --git a/flang/include/flang/Optimizer/Transforms/Passes.td b/flang/include/flang/Optimizer/Transforms/Passes.td
@@ -411,7 +411,10 @@ def FunctionAttr : Pass<"function-attr", "mlir::func::FuncOp"> {
     Option<"unsafeFPMath", "unsafe-fp-math",
            "bool", /*default=*/"false",
            "Set the unsafe-fp-math attribute on functions in the module.">,
-  ];
+    Option<"tuneCPU", "tune-cpu",
+           "llvm::StringRef", /*default=*/"llvm::StringRef{}",
+           "Set the tune-cpu attribute on functions in the module.">,
+];
 }
 
 def AssumedRankOpConversion : Pass<"fir-assumed-rank-op", "mlir::ModuleOp"> {
diff --git a/flang/lib/Frontend/CompilerInvocation.cpp b/flang/lib/Frontend/CompilerInvocation.cpp
@@ -431,6 +431,10 @@ static void parseTargetArgs(TargetOptions &opts, llvm::opt::ArgList &args) {
           args.getLastArg(clang::driver::options::OPT_target_cpu))
     opts.cpu = a->getValue();
 
+  if (const llvm::opt::Arg *a =
+          args.getLastArg(clang::driver::options::OPT_tune_cpu))
+    opts.cpuToTuneFor = a->getValue();
+
   for (const llvm::opt::Arg *currentArg :
        args.filtered(clang::driver::options::OPT_target_feature))
     opts.featuresAsWritten.emplace_back(currentArg->getValue());
diff --git a/flang/lib/Frontend/FrontendActions.cpp b/flang/lib/Frontend/FrontendActions.cpp
@@ -297,7 +297,8 @@ bool CodeGenAction::beginSourceFileAction() {
       ci.getParsing().allCooked(), ci.getInvocation().getTargetOpts().triple,
       kindMap, ci.getInvocation().getLoweringOpts(),
       ci.getInvocation().getFrontendOpts().envDefaults,
-      ci.getInvocation().getFrontendOpts().features, targetMachine);
+      ci.getInvocation().getFrontendOpts().features, targetMachine,
+      ci.getInvocation().getTargetOpts().cpuToTuneFor);
 
   // Fetch module from lb, so we can set
   mlirModule = std::make_unique<mlir::ModuleOp>(lb.getModule());
diff --git a/flang/lib/Lower/Bridge.cpp b/flang/lib/Lower/Bridge.cpp
@@ -6025,7 +6025,7 @@ Fortran::lower::LoweringBridge::LoweringBridge(
     const Fortran::lower::LoweringOptions &loweringOptions,
     const std::vector<Fortran::lower::EnvironmentDefault> &envDefaults,
     const Fortran::common::LanguageFeatureControl &languageFeatures,
-    const llvm::TargetMachine &targetMachine)
+    const llvm::TargetMachine &targetMachine, const llvm::StringRef tuneCPU)
     : semanticsContext{semanticsContext}, defaultKinds{defaultKinds},
       intrinsics{intrinsics}, targetCharacteristics{targetCharacteristics},
       cooked{&cooked}, context{context}, kindMap{kindMap},
@@ -6082,6 +6082,7 @@ Fortran::lower::LoweringBridge::LoweringBridge(
   fir::setTargetTriple(*module.get(), triple);
   fir::setKindMapping(*module.get(), kindMap);
   fir::setTargetCPU(*module.get(), targetMachine.getTargetCPU());
+  fir::setTuneCPU(*module.get(), tuneCPU);
   fir::setTargetFeatures(*module.get(), targetMachine.getTargetFeatureString());
   fir::support::setMLIRDataLayout(*module.get(),
                                   targetMachine.createDataLayout());
diff --git a/flang/lib/Optimizer/CodeGen/CodeGen.cpp b/flang/lib/Optimizer/CodeGen/CodeGen.cpp
@@ -3595,6 +3595,9 @@ class FIRToLLVMLowering
     if (!forcedTargetCPU.empty())
       fir::setTargetCPU(mod, forcedTargetCPU);
 
+    if (!forcedTuneCPU.empty())
+      fir::setTuneCPU(mod, forcedTuneCPU);
+
     if (!forcedTargetFeatures.empty())
       fir::setTargetFeatures(mod, forcedTargetFeatures);
 
diff --git a/flang/lib/Optimizer/CodeGen/Target.cpp b/flang/lib/Optimizer/CodeGen/Target.cpp
@@ -1113,3 +1113,14 @@ fir::CodeGenSpecifics::get(mlir::MLIRContext *ctx, llvm::Triple &&trp,
   }
   TODO(mlir::UnknownLoc::get(ctx), "target not implemented");
 }
+
+std::unique_ptr<fir::CodeGenSpecifics> fir::CodeGenSpecifics::get(
+    mlir::MLIRContext *ctx, llvm::Triple &&trp, KindMapping &&kindMap,
+    llvm::StringRef targetCPU, mlir::LLVM::TargetFeaturesAttr targetFeatures,
+    const mlir::DataLayout &dl, llvm::StringRef tuneCPU) {
+  std::unique_ptr<fir::CodeGenSpecifics> CGS = fir::CodeGenSpecifics::get(
+      ctx, std::move(trp), std::move(kindMap), targetCPU, targetFeatures, dl);
+
+  CGS->tuneCPU = tuneCPU;
+  return CGS;
+}
diff --git a/flang/lib/Optimizer/CodeGen/TargetRewrite.cpp b/flang/lib/Optimizer/CodeGen/TargetRewrite.cpp
@@ -89,6 +89,9 @@ class TargetRewrite : public fir::impl::TargetRewritePassBase<TargetRewrite> {
     if (!forcedTargetCPU.empty())
       fir::setTargetCPU(mod, forcedTargetCPU);
 
+    if (!forcedTuneCPU.empty())
+      fir::setTuneCPU(mod, forcedTuneCPU);
+
     if (!forcedTargetFeatures.empty())
       fir::setTargetFeatures(mod, forcedTargetFeatures);
 
@@ -106,7 +109,8 @@ class TargetRewrite : public fir::impl::TargetRewritePassBase<TargetRewrite> {
 
     auto specifics = fir::CodeGenSpecifics::get(
         mod.getContext(), fir::getTargetTriple(mod), fir::getKindMapping(mod),
-        fir::getTargetCPU(mod), fir::getTargetFeatures(mod), *dl);
+        fir::getTargetCPU(mod), fir::getTargetFeatures(mod), *dl,
+        fir::getTuneCPU(mod));
 
     setMembers(specifics.get(), &rewriter, &*dl);
 
@@ -672,12 +676,18 @@ class TargetRewrite : public fir::impl::TargetRewritePassBase<TargetRewrite> {
     auto targetCPU = specifics->getTargetCPU();
     mlir::StringAttr targetCPUAttr =
         targetCPU.empty() ? nullptr : mlir::StringAttr::get(ctx, targetCPU);
+    auto tuneCPU = specifics->getTuneCPU();
+    mlir::StringAttr tuneCPUAttr =
+        tuneCPU.empty() ? nullptr : mlir::StringAttr::get(ctx, tuneCPU);
     auto targetFeaturesAttr = specifics->getTargetFeatures();
 
     for (auto fn : mod.getOps<mlir::func::FuncOp>()) {
       if (targetCPUAttr)
         fn->setAttr("target_cpu", targetCPUAttr);
 
+      if (tuneCPUAttr)
+        fn->setAttr("tune_cpu", tuneCPUAttr);
+
       if (targetFeaturesAttr)
         fn->setAttr("target_features", targetFeaturesAttr);
 
diff --git a/flang/lib/Optimizer/CodeGen/TypeConverter.cpp b/flang/lib/Optimizer/CodeGen/TypeConverter.cpp
@@ -35,7 +35,8 @@ LLVMTypeConverter::LLVMTypeConverter(mlir::ModuleOp module, bool applyTBAA,
       kindMapping(getKindMapping(module)),
       specifics(CodeGenSpecifics::get(
           module.getContext(), getTargetTriple(module), getKindMapping(module),
-          getTargetCPU(module), getTargetFeatures(module), dl)),
+          getTargetCPU(module), getTargetFeatures(module), dl,
+          getTuneCPU(module))),
       tbaaBuilder(std::make_unique<TBAABuilder>(module->getContext(), applyTBAA,
                                                 forceUnifiedTBAATree)),
       dataLayout{&dl} {
diff --git a/flang/lib/Optimizer/Dialect/Support/FIRContext.cpp b/flang/lib/Optimizer/Dialect/Support/FIRContext.cpp
@@ -77,6 +77,24 @@ llvm::StringRef fir::getTargetCPU(mlir::ModuleOp mod) {
   return {};
 }
 
+static constexpr const char *tuneCpuName = "fir.tune_cpu";
+
+void fir::setTuneCPU(mlir::ModuleOp mod, llvm::StringRef cpu) {
+  if (cpu.empty())
+    return;
+
+  auto *ctx = mod.getContext();
+
+  mod->setAttr(tuneCpuName, mlir::StringAttr::get(ctx, cpu));
+}
+
+llvm::StringRef fir::getTuneCPU(mlir::ModuleOp mod) {
+  if (auto attr = mod->getAttrOfType<mlir::StringAttr>(tuneCpuName))
+    return attr.getValue();
+
+  return {};
+}
+
 static constexpr const char *targetFeaturesName = "fir.target_features";
 
 void fir::setTargetFeatures(mlir::ModuleOp mod, llvm::StringRef features) {
diff --git a/flang/test/Driver/tune-cpu-fir.f90 b/flang/test/Driver/tune-cpu-fir.f90
@@ -0,0 +1,25 @@
+! RUN: %if aarch64-registered-target %{ %flang_fc1 -emit-fir -triple aarch64-unknown-linux-gnu -target-cpu aarch64 %s -o - | FileCheck %s --check-prefixes=ALL,ARMCPU %}
+! RUN: %if aarch64-registered-target %{ %flang_fc1 -emit-fir -triple aarch64-unknown-linux-gnu -tune-cpu neoverse-n1 %s -o - | FileCheck %s --check-prefixes=ALL,ARMTUNE %}
+! RUN: %if aarch64-registered-target %{ %flang_fc1 -emit-fir -triple aarch64-unknown-linux-gnu -target-cpu aarch64 -tune-cpu neoverse-n1 %s -o - | FileCheck %s --check-prefixes=ALL,ARMBOTH %}
+
+! RUN: %if x86-registered-target %{ %flang_fc1 -emit-fir -triple x86_64-unknown-linux-gnu -target-cpu x86-64 %s -o - | FileCheck %s --check-prefixes=ALL,X86CPU %}
+! RUN: %if x86-registered-target %{ %flang_fc1 -emit-fir -triple x86_64-unknown-linux-gnu -tune-cpu pentium4 %s -o - | FileCheck %s --check-prefixes=ALL,X86TUNE %}
+! RUN: %if x86-registered-target %{ %flang_fc1 -emit-fir -triple x86_64-unknown-linux-gnu -target-cpu x86-64 -tune-cpu pentium4 %s -o - | FileCheck %s --check-prefixes=ALL,X86BOTH %}
+
+! ALL: module attributes {
+
+! ARMCPU-SAME:      fir.target_cpu = "aarch64"
+! ARMCPU-NOT:       fir.tune_cpu = "neoverse-n1"
+
+! ARMTUNE-SAME:     fir.tune_cpu = "neoverse-n1"
+
+! ARMBOTH-SAME: fir.target_cpu = "aarch64"
+! ARMBOTH-SAME: fir.tune_cpu = "neoverse-n1"  
+
+! X86CPU-SAME:      fir.target_cpu = "x86-64"
+! X86CPU-NOT:       fir.tune_cpu = "pentium4"
+
+! X86TUNE-SAME:     fir.tune_cpu = "pentium4"
+
+! X86BOTH-SAME: fir.target_cpu = "x86-64"
+! X86BOTH-SAME: fir.tune_cpu = "pentium4"
diff --git a/flang/test/Lower/tune-cpu-llvm.f90 b/flang/test/Lower/tune-cpu-llvm.f90
@@ -0,0 +1,8 @@
+! RUN: %if x86-registered-target %{ %flang -mtune=pentium4 -S -emit-llvm %s -o - | FileCheck %s --check-prefixes=ALL,CHECK-X86 %}
+! RUN: %if aarch64-registered-target %{ %flang -mtune=neoverse-n1 -S -emit-llvm %s -o - | FileCheck %s --check-prefixes=ALL,CHECK-ARM %}
+
+!ALL: attributes #{{[0-9]+}} = {
+!CHECK-X86-SAME: "tune-cpu"="pentium4"
+!CHECK-ARM-SAME: "tune-cpu"="neoverse-n1"
+subroutine a
+end subroutine a
diff --git a/flang/tools/bbc/bbc.cpp b/flang/tools/bbc/bbc.cpp
@@ -367,11 +367,12 @@ static llvm::LogicalResult convertFortranSourceToMLIR(
   loweringOptions.setLowerToHighLevelFIR(useHLFIR || emitHLFIR);
   loweringOptions.setNSWOnLoopVarInc(setNSW);
   std::vector<Fortran::lower::EnvironmentDefault> envDefaults = {};
+  constexpr const char *tuneCPU = "";
   auto burnside = Fortran::lower::LoweringBridge::create(
       ctx, semanticsContext, defKinds, semanticsContext.intrinsics(),
       semanticsContext.targetCharacteristics(), parsing.allCooked(),
       targetTriple, kindMap, loweringOptions, envDefaults,
-      semanticsContext.languageFeatures(), targetMachine);
+      semanticsContext.languageFeatures(), targetMachine, tuneCPU);
   mlir::ModuleOp mlirModule = burnside.getModule();
   if (enableOpenMP) {
     if (enableOpenMPGPU && !enableOpenMPDevice) {
diff --git a/flang/tools/tco/tco.cpp b/flang/tools/tco/tco.cpp
@@ -58,6 +58,9 @@ static cl::opt<std::string> targetTriple("target",
 static cl::opt<std::string>
     targetCPU("target-cpu", cl::desc("specify a target CPU"), cl::init(""));
 
+static cl::opt<std::string> tuneCPU("tune-cpu", cl::desc("specify a tune CPU"),
+                                    cl::init(""));
+
 static cl::opt<std::string>
     targetFeatures("target-features", cl::desc("specify the target features"),
                    cl::init(""));
@@ -113,6 +116,7 @@ compileFIR(const mlir::PassPipelineCLParser &passPipeline) {
   fir::setTargetTriple(*owningRef, targetTriple);
   fir::setKindMapping(*owningRef, kindMap);
   fir::setTargetCPU(*owningRef, targetCPU);
+  fir::setTuneCPU(*owningRef, tuneCPU);
   fir::setTargetFeatures(*owningRef, targetFeatures);
   // tco is a testing tool, so it will happily use the target independent
   // data layout if none is on the module.
diff --git a/flang/unittests/Optimizer/FIRContextTest.cpp b/flang/unittests/Optimizer/FIRContextTest.cpp
diff --git a/mlir/include/mlir/Dialect/LLVMIR/LLVMOps.td b/mlir/include/mlir/Dialect/LLVMIR/LLVMOps.td
diff --git a/mlir/lib/Target/LLVMIR/ModuleImport.cpp b/mlir/lib/Target/LLVMIR/ModuleImport.cpp
diff --git a/mlir/lib/Target/LLVMIR/ModuleTranslation.cpp b/mlir/lib/Target/LLVMIR/ModuleTranslation.cpp
diff --git a/mlir/test/Target/LLVMIR/Import/tune-cpu.ll b/mlir/test/Target/LLVMIR/Import/tune-cpu.ll
diff --git a/mlir/test/Target/LLVMIR/tune-cpu.mlir b/mlir/test/Target/LLVMIR/tune-cpu.mlir
