diff --git a/clang/tools/amdgpu-arch/AMDGPUArch.cpp b/clang/tools/amdgpu-arch/AMDGPUArch.cpp
--- a/clang/tools/amdgpu-arch/AMDGPUArch.cpp
+++ b/clang/tools/amdgpu-arch/AMDGPUArch.cpp
@@ -11,12 +11,25 @@
 //
 //===----------------------------------------------------------------------===//
 
+#include "clang/Basic/Version.h"
+#include "llvm/Support/CommandLine.h"
 #include "llvm/Support/DynamicLibrary.h"
 #include "llvm/Support/Error.h"
 #include <memory>
 #include <string>
 #include <vector>
 
+using namespace llvm;
+
+static cl::opt<bool> Help("h", cl::desc("Alias for -help"), cl::Hidden);
+
+// Mark all our options with this category.
+static cl::OptionCategory AMDGPUArchCategory("amdgpu-arch options");
+
+static void PrintVersion(raw_ostream &OS) {
+  OS << clang::getClangToolFullVersion("amdgpu-arch") << '\n';
+}
+
 #if DYNAMIC_HSA
 typedef enum {
   HSA_STATUS_SUCCESS = 0x0,
@@ -102,6 +115,16 @@
 }
 
 int main(int argc, char *argv[]) {
+  cl::HideUnrelatedOptions(AMDGPUArchCategory);
+
+  cl::SetVersionPrinter(PrintVersion);
+  cl::ParseCommandLineOptions(
+      argc, argv,
+      "A tool to detect the presence of AMDGPU devices on the system. \n\n"
+      "The tool will output each detected GPU architecture separated by a\n"
+      "newline character. If multiple GPUs of the same architecture are found\n"
+      "a string will be printed for each\n");
+
   // Attempt to load the HSA runtime.
   if (llvm::Error Err = loadHSA()) {
     logAllUnhandledErrors(std::move(Err), llvm::errs());
diff --git a/clang/tools/amdgpu-arch/CMakeLists.txt b/clang/tools/amdgpu-arch/CMakeLists.txt
--- a/clang/tools/amdgpu-arch/CMakeLists.txt
+++ b/clang/tools/amdgpu-arch/CMakeLists.txt
@@ -15,6 +15,7 @@
 if(hsa-runtime64_FOUND AND NOT (LLVM_BUILD_32_BITS OR CMAKE_SIZEOF_VOID_P EQUAL 4))
   set_target_properties(amdgpu-arch PROPERTIES INSTALL_RPATH_USE_LINK_PATH ON)
   target_link_libraries(amdgpu-arch PRIVATE hsa-runtime64::hsa-runtime64)
+  clang_target_link_libraries(amdgpu-arch PRIVATE clangBasic)
 else()
   target_compile_definitions(amdgpu-arch PRIVATE "DYNAMIC_HSA")
 endif()
diff --git a/clang/tools/nvptx-arch/CMakeLists.txt b/clang/tools/nvptx-arch/CMakeLists.txt
--- a/clang/tools/nvptx-arch/CMakeLists.txt
+++ b/clang/tools/nvptx-arch/CMakeLists.txt
@@ -14,6 +14,7 @@
 # If we found the CUDA library directly we just dynamically link against it.
 if(CUDAToolkit_FOUND AND NOT (LLVM_BUILD_32_BITS OR CMAKE_SIZEOF_VOID_P EQUAL 4))
   target_link_libraries(nvptx-arch PRIVATE CUDA::cuda_driver)
+  clang_target_link_libraries(nvptx-arch PRIVATE clangBasic)
 else()
   target_compile_definitions(nvptx-arch PRIVATE "DYNAMIC_CUDA")
 endif()
diff --git a/clang/tools/nvptx-arch/NVPTXArch.cpp b/clang/tools/nvptx-arch/NVPTXArch.cpp
--- a/clang/tools/nvptx-arch/NVPTXArch.cpp
+++ b/clang/tools/nvptx-arch/NVPTXArch.cpp
@@ -11,12 +11,25 @@
 //
 //===----------------------------------------------------------------------===//
 
+#include "clang/Basic/Version.h"
+#include "llvm/Support/CommandLine.h"
 #include "llvm/Support/DynamicLibrary.h"
 #include "llvm/Support/Error.h"
 #include <cstdint>
 #include <cstdio>
 #include <memory>
 
+using namespace llvm;
+
+static cl::opt<bool> Help("h", cl::desc("Alias for -help"), cl::Hidden);
+
+static void PrintVersion(raw_ostream &OS) {
+  OS << clang::getClangToolFullVersion("nvptx-arch") << '\n';
+}
+// Mark all our options with this category, everything else (except for -version
+// and -help) will be hidden.
+static cl::OptionCategory NVPTXArchCategory("nvptx-arch options");
+
 #if DYNAMIC_CUDA
 typedef enum cudaError_enum {
   CUDA_SUCCESS = 0,
@@ -79,6 +92,16 @@
 }
 
 int main(int argc, char *argv[]) {
+  cl::HideUnrelatedOptions(NVPTXArchCategory);
+
+  cl::SetVersionPrinter(PrintVersion);
+  cl::ParseCommandLineOptions(
+      argc, argv,
+      "A tool to detect the presence of NVIDIA devices on the system. \n\n"
+      "The tool will output each detected GPU architecture separated by a\n"
+      "newline character. If multiple GPUs of the same architecture are found\n"
+      "a string will be printed for each\n");
+
   // Attempt to load the NVPTX driver runtime.
   if (llvm::Error Err = loadCUDA()) {
     logAllUnhandledErrors(std::move(Err), llvm::errs());