Revert "[LinkerWrapper] Clean up options after proper forwarding" (#126495)

Reverts llvm/llvm-project#126297

Broken buildbots
https://lab.llvm.org/staging/#/builders/105/builds/15554
https://lab.llvm.org/buildbot/#/builders/30/builds/15490

Error is
```
# .---command stderr------------
# | FileCheck error: '/work/janplehr/git/llvm-project/bot-tester-builds/cmakecachebuild/runtimes/runtimes-bins/offload/test/amdgcn-amd-amdhsa/offloading/Output/bug51781.c.tmp.custom' is empty.
# | FileCheck command line:  /home/janplehr/git/llvm-project/bot-tester-builds/cmakecachebuild/./bin/FileCheck /work/janplehr/git/llvm-project/offload/test/offloading/bug51781.c -check-prefix=CUSTOM -input-file=/work/janplehr/git/llvm-project/bot-tester-builds/cmakecachebuild/runtimes/runtimes-bins/offload/test/amdgcn-amd-amdhsa/offloading/Output/bug51781.c.tmp.custom
```
The file is empty, while the `CUSTOM` check-target expects to find
```
// CUSTOM: Rewriting generic-mode kernel with a customized state machine.
```

clang/lib/Driver/ToolChains/Clang.cpp

@@ -9252,14 +9252,6 @@ void LinkerWrapper::ConstructJob(Compilation &C, const JobAction &JA,
       for (StringRef Arg : LinkerArgs)
         CmdArgs.push_back(Args.MakeArgString(
             "--device-linker=" + TC->getTripleString() + "=" + Arg));
-
-      // Forward the LTO mode relying on the Driver's parsing.
-      if (C.getDriver().getOffloadLTOMode() == LTOK_Full)
-        CmdArgs.push_back(Args.MakeArgString(
-            "--device-compiler=" + TC->getTripleString() + "=-flto=full"));
-      else if (C.getDriver().getOffloadLTOMode() == LTOK_Thin)
-        CmdArgs.push_back(Args.MakeArgString(
-            "--device-compiler=" + TC->getTripleString() + "=-flto=thin"));
     }
   }
 
@@ -9267,9 +9259,6 @@ void LinkerWrapper::ConstructJob(Compilation &C, const JobAction &JA,
       Args.MakeArgString("--host-triple=" + getToolChain().getTripleString()));
   if (Args.hasArg(options::OPT_v))
     CmdArgs.push_back("--wrapper-verbose");
-  if (Arg *A = Args.getLastArg(options::OPT_cuda_path_EQ))
-    CmdArgs.push_back(
-        Args.MakeArgString(Twine("--cuda-path=") + A->getValue()));
 
   // Construct the link job so we can wrap around it.
   Linker->ConstructJob(C, JA, Output, Inputs, Args, LinkingOutput);

clang/test/Driver/linker-wrapper.c

@@ -21,16 +21,16 @@ __attribute__((visibility("protected"), used)) int x;
 // RUN: clang-linker-wrapper --host-triple=x86_64-unknown-linux-gnu --dry-run \
 // RUN:   --linker-path=/usr/bin/ld %t.o -o a.out 2>&1 | FileCheck %s --check-prefix=NVPTX-LINK
 
-// NVPTX-LINK: clang{{.*}} -o {{.*}}.img --target=nvptx64-nvidia-cuda -march=sm_70 {{.*}}.o {{.*}}.o
+// NVPTX-LINK: clang{{.*}} -o {{.*}}.img --target=nvptx64-nvidia-cuda -march=sm_70 -O2 -flto {{.*}}.o {{.*}}.o
 
 // RUN: clang-offload-packager -o %t.out \
 // RUN:   --image=file=%t.elf.o,kind=openmp,triple=nvptx64-nvidia-cuda,arch=sm_70 \
 // RUN:   --image=file=%t.elf.o,kind=openmp,triple=nvptx64-nvidia-cuda,arch=sm_70
 // RUN: %clang -cc1 %s -triple x86_64-unknown-linux-gnu -emit-obj -o %t.o -fembed-offload-object=%t.out
-// RUN: clang-linker-wrapper --host-triple=x86_64-unknown-linux-gnu --dry-run --device-compiler=-g \
+// RUN: clang-linker-wrapper --host-triple=x86_64-unknown-linux-gnu --dry-run --device-debug -O0 \
 // RUN:   --linker-path=/usr/bin/ld %t.o -o a.out 2>&1 | FileCheck %s --check-prefix=NVPTX-LINK-DEBUG
 
-// NVPTX-LINK-DEBUG: clang{{.*}} --target=nvptx64-nvidia-cuda -march=sm_70 {{.*}}-g
+// NVPTX-LINK-DEBUG: clang{{.*}} -o {{.*}}.img --target=nvptx64-nvidia-cuda -march=sm_70 -O2 -flto {{.*}}.o {{.*}}.o -g
 
 // RUN: clang-offload-packager -o %t.out \
 // RUN:   --image=file=%t.elf.o,kind=openmp,triple=amdgcn-amd-amdhsa,arch=gfx908 \
@@ -39,16 +39,16 @@ __attribute__((visibility("protected"), used)) int x;
 // RUN: clang-linker-wrapper --host-triple=x86_64-unknown-linux-gnu --dry-run \
 // RUN:   --linker-path=/usr/bin/ld %t.o -o a.out 2>&1 | FileCheck %s --check-prefix=AMDGPU-LINK
 
-// AMDGPU-LINK: clang{{.*}} -o {{.*}}.img --target=amdgcn-amd-amdhsa -mcpu=gfx908 -Wl,--no-undefined {{.*}}.o {{.*}}.o
+// AMDGPU-LINK: clang{{.*}} -o {{.*}}.img --target=amdgcn-amd-amdhsa -mcpu=gfx908 -O2 -flto -Wl,--no-undefined {{.*}}.o {{.*}}.o
 
 // RUN: clang-offload-packager -o %t.out \
 // RUN:   --image=file=%t.amdgpu.bc,kind=openmp,triple=amdgcn-amd-amdhsa,arch=gfx1030 \
 // RUN:   --image=file=%t.amdgpu.bc,kind=openmp,triple=amdgcn-amd-amdhsa,arch=gfx1030
 // RUN: %clang -cc1 %s -triple x86_64-unknown-linux-gnu -emit-obj -o %t.o -fembed-offload-object=%t.out
-// RUN: clang-linker-wrapper --host-triple=x86_64-unknown-linux-gnu --dry-run --device-compiler=--save-temps \
+// RUN: clang-linker-wrapper --host-triple=x86_64-unknown-linux-gnu --dry-run --save-temps -O2 \
 // RUN:   --linker-path=/usr/bin/ld %t.o -o a.out 2>&1 | FileCheck %s --check-prefix=AMDGPU-LTO-TEMPS
 
-// AMDGPU-LTO-TEMPS: clang{{.*}} --target=amdgcn-amd-amdhsa -mcpu=gfx1030 {{.*}}-save-temps
+// AMDGPU-LTO-TEMPS: clang{{.*}} -o {{.*}}.img --target=amdgcn-amd-amdhsa -mcpu=gfx1030 -O2 -flto -Wl,--no-undefined {{.*}}.o -save-temps
 
 // RUN: clang-offload-packager -o %t.out \
 // RUN:   --image=file=%t.elf.o,kind=openmp,triple=x86_64-unknown-linux-gnu \
@@ -59,7 +59,7 @@ __attribute__((visibility("protected"), used)) int x;
 // RUN:   --linker-path=/usr/bin/ld.lld --whole-archive %t.a --no-whole-archive \
 // RUN:   %t.o -o a.out 2>&1 | FileCheck %s --check-prefix=CPU-LINK
 
-// CPU-LINK: clang{{.*}} -o {{.*}}.img --target=x86_64-unknown-linux-gnu -march=native -Wl,--no-undefined {{.*}}.o {{.*}}.o -Wl,-Bsymbolic -shared -Wl,--whole-archive {{.*}}.a -Wl,--no-whole-archive
+// CPU-LINK: clang{{.*}} -o {{.*}}.img --target=x86_64-unknown-linux-gnu -march=native -O2 -flto -Wl,--no-undefined {{.*}}.o {{.*}}.o -Wl,-Bsymbolic -shared -Wl,--whole-archive {{.*}}.a -Wl,--no-whole-archive
 
 // RUN: %clang -cc1 %s -triple x86_64-unknown-linux-gnu -emit-obj -o %t.o
 // RUN: clang-linker-wrapper --dry-run --host-triple=x86_64-unknown-linux-gnu -mllvm -openmp-opt-disable \
@@ -148,7 +148,7 @@ __attribute__((visibility("protected"), used)) int x;
 // RUN: clang-linker-wrapper --host-triple=x86_64-unknown-linux-gnu --dry-run --clang-backend \
 // RUN:   --linker-path=/usr/bin/ld %t.o -o a.out 2>&1 | FileCheck %s --check-prefix=CLANG-BACKEND
 
-// CLANG-BACKEND: clang{{.*}} -o {{.*}}.img --target=amdgcn-amd-amdhsa -mcpu=gfx908 -Wl,--no-undefined {{.*}}.o
+// CLANG-BACKEND: clang{{.*}} -o {{.*}}.img --target=amdgcn-amd-amdhsa -mcpu=gfx908 -O2 -flto -Wl,--no-undefined {{.*}}.o
 
 // RUN: clang-offload-packager -o %t.out \
 // RUN:   --image=file=%t.elf.o,kind=openmp,triple=nvptx64-nvidia-cuda,arch=sm_70
@@ -171,8 +171,8 @@ __attribute__((visibility("protected"), used)) int x;
 // RUN: clang-linker-wrapper --host-triple=x86_64-unknown-linux-gnu --dry-run \
 // RUN:   --linker-path=/usr/bin/ld %t-on.o %t-off.o %t.a -o a.out 2>&1 | FileCheck %s --check-prefix=AMD-TARGET-ID
 
-// AMD-TARGET-ID: clang{{.*}} -o {{.*}}.img --target=amdgcn-amd-amdhsa -mcpu=gfx90a:xnack+ -Wl,--no-undefined {{.*}}.o {{.*}}.o
-// AMD-TARGET-ID: clang{{.*}} -o {{.*}}.img --target=amdgcn-amd-amdhsa -mcpu=gfx90a:xnack- -Wl,--no-undefined {{.*}}.o {{.*}}.o
+// AMD-TARGET-ID: clang{{.*}} -o {{.*}}.img --target=amdgcn-amd-amdhsa -mcpu=gfx90a:xnack+ -O2 -flto -Wl,--no-undefined {{.*}}.o {{.*}}.o
+// AMD-TARGET-ID: clang{{.*}} -o {{.*}}.img --target=amdgcn-amd-amdhsa -mcpu=gfx90a:xnack- -O2 -flto -Wl,--no-undefined {{.*}}.o {{.*}}.o
 
 // RUN: clang-offload-packager -o %t-lib.out \
 // RUN:   --image=file=%t.elf.o,kind=openmp,triple=amdgcn-amd-amdhsa,arch=generic
@@ -187,8 +187,8 @@ __attribute__((visibility("protected"), used)) int x;
 // RUN: clang-linker-wrapper --host-triple=x86_64-unknown-linux-gnu --dry-run \
 // RUN:   --linker-path=/usr/bin/ld %t1.o %t2.o %t.a -o a.out 2>&1 | FileCheck %s --check-prefix=ARCH-ALL
 
-// ARCH-ALL: clang{{.*}} -o {{.*}}.img --target=amdgcn-amd-amdhsa -mcpu=gfx90a -Wl,--no-undefined {{.*}}.o {{.*}}.o
-// ARCH-ALL: clang{{.*}} -o {{.*}}.img --target=amdgcn-amd-amdhsa -mcpu=gfx908 -Wl,--no-undefined {{.*}}.o {{.*}}.o
+// ARCH-ALL: clang{{.*}} -o {{.*}}.img --target=amdgcn-amd-amdhsa -mcpu=gfx90a -O2 -flto -Wl,--no-undefined {{.*}}.o {{.*}}.o
+// ARCH-ALL: clang{{.*}} -o {{.*}}.img --target=amdgcn-amd-amdhsa -mcpu=gfx908 -O2 -flto -Wl,--no-undefined {{.*}}.o {{.*}}.o
 
 // RUN: clang-offload-packager -o %t.out \
 // RUN:   --image=file=%t.elf.o,kind=openmp,triple=x86_64-unknown-linux-gnu \

clang/test/Driver/openmp-offload.c

@@ -208,13 +208,3 @@
 // RUN:     -fsyntax-only %s 2>&1 | FileCheck -check-prefix=CHK-SYNTAX-ONLY-ARGS %s
 // CHK-SYNTAX-ONLY-ARGS: "-cc1" "-triple" "powerpc64le-ibm-linux-gnu"{{.*}}"-fsyntax-only"
 // CHK-SYNTAX-ONLY-ARGS: "-cc1" "-triple" "powerpc64le-unknown-linux"{{.*}}"-fsyntax-only"
-
-//
-// Ensure `-foffload-lto` is forwarded properly.
-//
-// RUN:   %clang -### --target=powerpc64le-linux -fopenmp=libomp -fopenmp-targets=powerpc64le-ibm-linux-gnu \
-// RUN:     -foffload-lto %s 2>&1 | FileCheck -check-prefix=CHK-DEVICE-LTO-FULL %s
-// CHK-DEVICE-LTO-FULL: clang-linker-wrapper{{.*}} "--device-compiler=powerpc64le-ibm-linux-gnu=-flto=full"
-// RUN:   %clang -### --target=powerpc64le-linux -fopenmp=libomp -fopenmp-targets=powerpc64le-ibm-linux-gnu \
-// RUN:     -foffload-lto=thin %s 2>&1 | FileCheck -check-prefix=CHK-DEVICE-LTO-THIN %s
-// CHK-DEVICE-LTO-THIN: clang-linker-wrapper{{.*}} "--device-compiler=powerpc64le-ibm-linux-gnu=-flto=thin"

clang/tools/clang-linker-wrapper/ClangLinkerWrapper.cpp

@@ -485,6 +485,7 @@ Expected<StringRef> clang(ArrayRef<StringRef> InputFiles, const ArgList &Args) {
   if (!TempFileOrErr)
     return TempFileOrErr.takeError();
 
+  StringRef OptLevel = Args.getLastArgValue(OPT_opt_level, "O2");
   SmallVector<StringRef, 16> CmdArgs{
       *ClangPath,
       "--no-default-config",
@@ -492,9 +493,12 @@ Expected<StringRef> clang(ArrayRef<StringRef> InputFiles, const ArgList &Args) {
       *TempFileOrErr,
       Args.MakeArgString("--target=" + Triple.getTriple()),
       Triple.isAMDGPU() ? Args.MakeArgString("-mcpu=" + Arch)
-                        : Args.MakeArgString("-march=" + Arch)};
+                        : Args.MakeArgString("-march=" + Arch),
+      Args.MakeArgString("-" + OptLevel),
+  };
 
   // Forward all of the `--offload-opt` and similar options to the device.
+  CmdArgs.push_back("-flto");
   for (auto &Arg : Args.filtered(OPT_offload_opt_eq_minus, OPT_mllvm))
     CmdArgs.append(
         {"-Xlinker",
@@ -543,12 +547,29 @@ Expected<StringRef> clang(ArrayRef<StringRef> InputFiles, const ArgList &Args) {
     CmdArgs.append({"-Xlinker", Args.MakeArgString(
                                     "-mllvm=" + StringRef(Arg->getValue()))});
 
+  if (Args.hasArg(OPT_debug))
+    CmdArgs.push_back("-g");
+
+  if (SaveTemps)
+    CmdArgs.push_back("-save-temps");
+
   if (SaveTemps && linkerSupportsLTO(Args))
     CmdArgs.push_back("-Wl,--save-temps");
 
   if (Args.hasArg(OPT_embed_bitcode))
     CmdArgs.push_back("-Wl,--lto-emit-llvm");
 
+  if (Verbose)
+    CmdArgs.push_back("-v");
+
+  if (!CudaBinaryPath.empty())
+    CmdArgs.push_back(Args.MakeArgString("--cuda-path=" + CudaBinaryPath));
+
+  for (StringRef Arg : Args.getAllArgValues(OPT_ptxas_arg))
+    llvm::copy(
+        SmallVector<StringRef>({"-Xcuda-ptxas", Args.MakeArgString(Arg)}),
+        std::back_inserter(CmdArgs));
+
   for (StringRef Arg : Args.getAllArgValues(OPT_linker_arg_EQ))
     CmdArgs.append({"-Xlinker", Args.MakeArgString(Arg)});
   for (StringRef Arg : Args.getAllArgValues(OPT_compiler_arg_EQ))

clang/tools/clang-linker-wrapper/LinkerWrapperOpts.td

@@ -17,9 +17,11 @@ def cuda_path_EQ : Joined<["--"], "cuda-path=">,
   Flags<[WrapperOnlyOption]>, MetaVarName<"<dir>">,
   HelpText<"Set the system CUDA path">;
 def host_triple_EQ : Joined<["--"], "host-triple=">,
-                     Flags<[WrapperOnlyOption]>,
-                     MetaVarName<"<triple>">,
-                     HelpText<"Triple to use for the host compilation">;
+  Flags<[WrapperOnlyOption]>, MetaVarName<"<triple>">,
+  HelpText<"Triple to use for the host compilation">;
+def opt_level : Joined<["--"], "opt-level=">,
+  Flags<[WrapperOnlyOption]>, MetaVarName<"<O0, O1, O2, or O3>">,
+  HelpText<"Optimization level for LTO">;
 def device_linker_args_EQ : Joined<["--"], "device-linker=">,
   Flags<[WrapperOnlyOption]>, MetaVarName<"<value> or <triple>=<value>">,
   HelpText<"Arguments to pass to the device linker invocation">;
@@ -32,8 +34,18 @@ def dry_run : Flag<["--"], "dry-run">,
 def verbose : Flag<["--"], "wrapper-verbose">,
   Flags<[WrapperOnlyOption]>, HelpText<"Verbose output from tools">;
 def embed_bitcode : Flag<["--"], "embed-bitcode">,
-                    Flags<[WrapperOnlyOption]>,
-                    HelpText<"Embed linked bitcode in the module">;
+  Flags<[WrapperOnlyOption]>, HelpText<"Embed linked bitcode in the module">;
+def debug : Flag<["--"], "device-debug">, Flags<[WrapperOnlyOption]>, 
+  HelpText<"Use debugging">;
+def ptxas_arg : Joined<["--"], "ptxas-arg=">,
+  Flags<[WrapperOnlyOption]>,
+  HelpText<"Argument to pass to the 'ptxas' invocation">;
+def pass_remarks_EQ : Joined<["--"], "pass-remarks=">,
+  Flags<[WrapperOnlyOption]>, HelpText<"Pass remarks for LTO">;
+def pass_remarks_missed_EQ : Joined<["--"], "pass-remarks-missed=">,
+  Flags<[WrapperOnlyOption]>, HelpText<"Pass remarks for LTO">;
+def pass_remarks_analysis_EQ : Joined<["--"], "pass-remarks-analysis=">,
+  Flags<[WrapperOnlyOption]>, HelpText<"Pass remarks for LTO">;
 def print_wrapped_module : Flag<["--"], "print-wrapped-module">,
   Flags<[WrapperOnlyOption]>,
   HelpText<"Print the wrapped module's IR for testing">;