[PATCH] D119841: [OpenMP] Pass AMDGPU math libraries into the linker wrapper

[Joseph Huber via Phabricator via cfe-commits]

jhuber6 created this revision.
jhuber6 added reviewers: JonChesterfield, jdoerfert.
Herald added subscribers: kerbowa, guansong, t-tye, tpr, dstuttard, yaxunl, jvesely, kzhuravl.
jhuber6 requested review of this revision.
Herald added subscribers: cfe-commits, sstefan1, wdng.
Herald added a project: clang.

This patch passes in the AMDPGU math libraries to the linker wrapper.
The wrapper already handles linking OpenMP bitcode libraries via the
`--target-library` option. This should be sufficient to link in math
libraries for the accompanying architecture.


Repository:
  rG LLVM Github Monorepo

https://reviews.llvm.org/D119841

Files:
  clang/lib/Driver/ToolChains/Clang.cpp
  clang/test/Driver/amdgpu-openmp-toolchain.c


Index: clang/test/Driver/amdgpu-openmp-toolchain.c
===================================================================
--- clang/test/Driver/amdgpu-openmp-toolchain.c
+++ clang/test/Driver/amdgpu-openmp-toolchain.c
@@ -77,3 +77,6 @@
 
 // RUN: %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: {{.*}}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"
+
+// RUN: %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 -fopenmp-new-driver %s 2>&1 | FileCheck %s --check-prefix=CHECK-LIB-DEVICE-NEW
+// CHECK-LIB-DEVICE-NEW: {{.*}}clang-linker-wrapper{{.*}}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"
Index: clang/lib/Driver/ToolChains/Clang.cpp
===================================================================
--- clang/lib/Driver/ToolChains/Clang.cpp
+++ clang/lib/Driver/ToolChains/Clang.cpp
@@ -8191,6 +8191,33 @@
     }
   }
 
+  // Get the AMDGPU math libraries.
+  // FIXME: This method is bad, remove once AMDGPU has a proper math library.
+  for (auto &I : llvm::make_range(OpenMPTCRange.first, OpenMPTCRange.second)) {
+    const ToolChain *TC = I.second;
+
+    if (!TC->getTriple().isAMDGPU())
+      continue;
+
+    // Don't pass math libraries if this is C and it wasn't specified.
+    const auto LibraryArgs = Args.getAllArgValues(options::OPT_l);
+    if (llvm::find(LibraryArgs, "m") == LibraryArgs.end() && !D.CCCIsCXX())
+      continue;
+
+    const ArgList &TCArgs = C.getArgsForToolChain(TC, "", Action::OFK_OpenMP);
+    StringRef Arch = TCArgs.getLastArgValue(options::OPT_march_EQ);
+    const toolchains::ROCMToolChain RocmTC(TC->getDriver(), TC->getTriple(),
+                                           TCArgs);
+
+    SmallVector<std::string, 12> BCLibs =
+        RocmTC.getCommonDeviceLibNames(TCArgs, Arch.str());
+
+    for (StringRef LibName : BCLibs)
+      CmdArgs.push_back(
+          Args.MakeArgString("-target-library=" + TC->getTripleString() + "-" +
+                             Arch + "=" + LibName));
+  }
+
   if (D.isUsingLTO(/* IsOffload */ true)) {
     // Pass in target features for each toolchain.
     for (auto &I :


-------------- next part --------------
A non-text attachment was scrubbed...
Name: D119841.408856.patch
Type: text/x-patch
Size: 2764 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/cfe-commits/attachments/20220215/a7dd7da4/attachment.bin>