diff --git a/mlir/CMakeLists.txt b/mlir/CMakeLists.txt
--- a/mlir/CMakeLists.txt
+++ b/mlir/CMakeLists.txt
@@ -44,6 +44,8 @@
 else()
   set(MLIR_CUDA_CONVERSIONS_ENABLED 0)
 endif()
+# TODO: we should use a config.h file like LLVM does
+add_definitions(-DMLIR_CUDA_CONVERSIONS_ENABLED=${MLIR_CUDA_CONVERSIONS_ENABLED})
 
 set(MLIR_CUDA_RUNNER_ENABLED 0 CACHE BOOL "Enable building the mlir CUDA runner")
 
diff --git a/mlir/include/mlir/InitAllPasses.h b/mlir/include/mlir/InitAllPasses.h
--- a/mlir/include/mlir/InitAllPasses.h
+++ b/mlir/include/mlir/InitAllPasses.h
@@ -90,8 +90,10 @@
 
   // CUDA
   createConvertGpuLaunchFuncToCudaCallsPass();
+#if MLIR_CUDA_CONVERSIONS_ENABLED
   createConvertGPUKernelToCubinPass(
       [](const std::string &, Location, StringRef) { return nullptr; });
+#endif
   createLowerGpuOpsToNVVMOpsPass();
 
   // Linalg
diff --git a/mlir/lib/Conversion/GPUToCUDA/CMakeLists.txt b/mlir/lib/Conversion/GPUToCUDA/CMakeLists.txt
--- a/mlir/lib/Conversion/GPUToCUDA/CMakeLists.txt
+++ b/mlir/lib/Conversion/GPUToCUDA/CMakeLists.txt
@@ -1,7 +1,16 @@
-add_llvm_library(MLIRGPUtoCUDATransforms
+set(LLVM_OPTIONAL_SOURCES
   ConvertKernelFuncToCubin.cpp
+)
+
+set(SOURCES
   ConvertLaunchFuncToCudaCalls.cpp
 )
+
+if (MLIR_CUDA_CONVERSIONS_ENABLED)
+ list(APPEND SOURCES "ConvertKernelFuncToCubin.cpp")
+endif()
+
+add_llvm_library(MLIRGPUtoCUDATransforms ${SOURCES})
 target_link_libraries(MLIRGPUtoCUDATransforms
   MLIRGPU
   MLIRLLVMIR
diff --git a/mlir/lib/Conversion/GPUToCUDA/ConvertKernelFuncToCubin.cpp b/mlir/lib/Conversion/GPUToCUDA/ConvertKernelFuncToCubin.cpp
--- a/mlir/lib/Conversion/GPUToCUDA/ConvertKernelFuncToCubin.cpp
+++ b/mlir/lib/Conversion/GPUToCUDA/ConvertKernelFuncToCubin.cpp
@@ -57,9 +57,10 @@
     gpu::GPUModuleOp module = getOperation();
 
     // Make sure the NVPTX target is initialized.
-    llvm::InitializeAllTargets();
-    llvm::InitializeAllTargetMCs();
-    llvm::InitializeAllAsmPrinters();
+    LLVMInitializeNVPTXTarget();
+    LLVMInitializeNVPTXTargetInfo();
+    LLVMInitializeNVPTXTargetMC();
+    LLVMInitializeNVPTXAsmPrinter();
 
     auto llvmModule = translateModuleToNVVMIR(module);
     if (!llvmModule)