[AMDGPU][OpenMP] Use llvm-link to link ocml libraries

[pdhaliwal]

Summary

This fixes the 'unused linker option: -lm' warning when compiling
program with -c.

[JonChesterfield]

This is semantically different to using mlink-builtin-bitcode with clang. The former internalizes all the symbols after they are introduced, using llvm-link does not. I'm not immediately sure whether that is a problem - the symbols in rocm device libs should all be __prefixed - but it's potentially bad for code size.

Suggest we fix the test case and leave restoring internalize for a later patch, thoughts?

clang/test/Driver/amdgpu-openmp-toolchain.c

80

This will fail on machines with a different directory layout, perhaps copy the {{.*}} pattern from above

[pdhaliwal]

Remove redundant test

[pdhaliwal]

clang/test/Driver/amdgpu-openmp-toolchain.c

80

This was accidently put up as a comment to check the actual 'CHECK' line.

[JonChesterfield]

LG, thanks!

clang/lib/Driver/ToolChains/AMDGPUOpenMP.h

 
 #include "AMDGPU.h"
 #include "clang/Driver/Tool.h"
 #include "clang/Driver/ToolChain.h"
 
 namespace clang {
 namespace driver {
 
+namespace toolchains {
+class AMDGPUOpenMPToolChain;
+}
+
 namespace tools {
 
 namespace AMDGCN {
 // Runs llvm-link/opt/llc/lld, which links multiple LLVM bitcode, together with
 // device library, then compiles it to ISA in a shared object.
 class LLVM_LIBRARY_VISIBILITY OpenMPLinker : public Tool {
 public:
   OpenMPLinker(const ToolChain &TC)
       : Tool("AMDGCN::OpenMPLinker", "amdgcn-link", TC) {}
 
   bool hasIntegratedCPP() const override { return false; }
 
   void ConstructJob(Compilation &C, const JobAction &JA,
                     const InputInfo &Output, const InputInfoList &Inputs,
                     const llvm::opt::ArgList &TCArgs,
                     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;
+  const char *constructLLVMLinkCommand(
+      const toolchains::AMDGPUOpenMPToolChain &AMDGPUOpenMPTC, Compilation &C,
+      const JobAction &JA, const InputInfoList &Inputs,
+      const llvm::opt::ArgList &Args, llvm::StringRef SubArchName,
 
   /// \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,

clang/lib/Driver/ToolChains/AMDGPUOpenMP.cpp

if (auto Err = TC.getSystemGPUArch(Args, GPUArch)) {
     return false;
   }
 
   return true;
 }
 } // namespace
 
 const char *AMDGCN::OpenMPLinker::constructLLVMLinkCommand(
-    Compilation &C, const JobAction &JA, const InputInfoList &Inputs,
-    const ArgList &Args, StringRef SubArchName,
-    StringRef OutputFilePrefix) const {
+    const toolchains::AMDGPUOpenMPToolChain &AMDGPUOpenMPTC, Compilation &C,
+    const JobAction &JA, const InputInfoList &Inputs, const ArgList &Args,
+    StringRef SubArchName, StringRef OutputFilePrefix) const {
   ArgStringList CmdArgs;
 
   for (const auto &II : Inputs)
     if (II.isFilename())
       CmdArgs.push_back(II.getFilename());
+
+  if (Args.hasArg(options::OPT_l)) {
+    auto Lm = Args.getAllArgValues(options::OPT_l);
+    bool HasLibm = false;
+    for (auto &Lib : Lm) {
+      if (Lib == "m") {
+        HasLibm = true;
+        break;
+      }
+    }
+
+    if (HasLibm) {
+      SmallVector<std::string, 12> BCLibs =
+          AMDGPUOpenMPTC.getCommonDeviceLibNames(Args, SubArchName.str());
+      llvm::for_each(BCLibs, [&](StringRef BCFile) {
+        CmdArgs.push_back(Args.MakeArgString(BCFile));
+      });
+    }
+  }
+
   // Add an intermediate output file.
   CmdArgs.push_back("-o");
   const char *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>(
void AMDGCN::OpenMPLinker::ConstructJob(Compilation &C, const JobAction &JA,
   // Prefix for temporary file name.
   std::string Prefix;
   for (const auto &II : Inputs)
     if (II.isFilename())
       Prefix = llvm::sys::path::stem(II.getFilename()).str() + "-" + GPUArch;
   assert(Prefix.length() && "no linker inputs are files ");
 
   // Each command outputs different files.
-  const char *LLVMLinkCommand =
-      constructLLVMLinkCommand(C, JA, Inputs, Args, GPUArch, Prefix);
+  const char *LLVMLinkCommand = constructLLVMLinkCommand(
+      AMDGPUOpenMPTC, C, JA, Inputs, Args, GPUArch, Prefix);
 
   // Produce readable assembly if save-temps is enabled.
   if (C.getDriver().isSaveTempsEnabled())
     constructLlcCommand(C, JA, Inputs, Args, GPUArch, Prefix, LLVMLinkCommand,
                         /*OutputIsAsm=*/true);
   const char *LlcCommand = constructLlcCommand(C, JA, Inputs, Args, GPUArch,
                                                Prefix, LLVMLinkCommand);
   constructLldCommand(C, JA, Inputs, Output, Args, LlcCommand);
void AMDGPUOpenMPToolChain::addClangTargetOptions(
   if (DriverArgs.hasFlag(options::OPT_fopenmp_target_new_runtime,
                          options::OPT_fno_openmp_target_new_runtime, false))
     BitcodeSuffix = "new-amdgcn-" + GPUArch;
   else
     BitcodeSuffix = "amdgcn-" + GPUArch;
 
   addOpenMPDeviceRTL(getDriver(), DriverArgs, CC1Args, BitcodeSuffix,
                      getTriple());
-
-  if (!DriverArgs.hasArg(options::OPT_l))
-    return;
-
-  auto Lm = DriverArgs.getAllArgValues(options::OPT_l);
-  bool HasLibm = false;
-  for (auto &Lib : Lm) {
-    if (Lib == "m") {
-      HasLibm = true;
-      break;
-    }
-  }
-
-  if (HasLibm) {
-    SmallVector<std::string, 12> BCLibs =
-        getCommonDeviceLibNames(DriverArgs, GPUArch);
-    llvm::for_each(BCLibs, [&](StringRef BCFile) {
-      CC1Args.push_back("-mlink-builtin-bitcode");
-      CC1Args.push_back(DriverArgs.MakeArgString(BCFile));
-    });
-  }
 }
 
 llvm::opt::DerivedArgList *AMDGPUOpenMPToolChain::TranslateArgs(
     const llvm::opt::DerivedArgList &Args, StringRef BoundArch,
     Action::OffloadKind DeviceOffloadKind) const {
   DerivedArgList *DAL =
       HostTC.TranslateArgs(Args, BoundArch, DeviceOffloadKind);
   if (!DAL)

clang/test/Driver/amdgpu-openmp-toolchain.c

 // CHECK-C: "x86_64-unknown-linux-gnu" - "clang"
 // CHECK-C: "x86_64-unknown-linux-gnu" - "clang::as"
 // CHECK-C: "x86_64-unknown-linux-gnu" - "offload bundler"
 
 // RUN:   %clang -### --target=x86_64-unknown-linux-gnu -emit-llvm -S -fopenmp -fopenmp-targets=amdgcn-amd-amdhsa -Xopenmp-target=amdgcn-amd-amdhsa -march=gfx803 -nogpulib %s 2>&1 | FileCheck %s --check-prefix=CHECK-EMIT-LLVM-IR
 // CHECK-EMIT-LLVM-IR: clang{{.*}}"-cc1"{{.*}}"-triple" "amdgcn-amd-amdhsa"{{.*}}"-emit-llvm"
 
 // RUN: env LIBRARY_PATH=%S/Inputs/hip_dev_lib %clang -### -target x86_64-pc-linux-gnu -fopenmp -fopenmp-targets=amdgcn-amd-amdhsa -Xopenmp-target=amdgcn-amd-amdhsa -march=gfx803 -lm --rocm-device-lib-path=%S/Inputs/rocm/amdgcn/bitcode %s 2>&1 | FileCheck %s --check-prefix=CHECK-LIB-DEVICE
-// CHECK-LIB-DEVICE: clang{{.*}}"-cc1"{{.*}}"-triple" "amdgcn-amd-amdhsa"{{.*}}"-mlink-builtin-bitcode"{{.*}}libomptarget-amdgcn-gfx803.bc"{{.*}}"-mlink-builtin-bitcode"{{.*}}ocml.bc" "-mlink-builtin-bitcode"{{.*}}ockl.bc" "-mlink-builtin-bitcode"{{.*}}oclc_daz_opt_on.bc" "-mlink-builtin-bitcode"{{.*}}oclc_unsafe_math_off.bc" "-mlink-builtin-bitcode"{{.*}}oclc_finite_only_off.bc" "-mlink-builtin-bitcode"{{.*}}oclc_correctly_rounded_sqrt_on.bc" "-mlink-builtin-bitcode"{{.*}}oclc_wavefrontsize64_on.bc" "-mlink-builtin-bitcode"{{.*}}oclc_isa_version_803.bc"
+// CHECK-LIB-DEVICE: {{.*}}llvm-link{{.*}}ocml.bc"{{.*}}ockl.bc"{{.*}}oclc_daz_opt_on.bc"{{.*}}oclc_unsafe_math_off.bc"{{.*}}oclc_finite_only_off.bc"{{.*}}oclc_correctly_rounded_sqrt_on.bc"{{.*}}oclc_wavefrontsize64_on.bc"{{.*}}oclc_isa_version_803.bc"