[mlir][gpu] Add missing initialization of gpu runtime wrappers.

mlir/tools/mlir-cuda-runner/cuda-runtime-wrappers.cpp

[](CUresult result) {                                                        \
     cuGetErrorName(result, &name);                                             \
     if (!name)                                                                 \
       name = "<unknown>";                                                      \
     llvm::errs() << "'" << #expr << "' failed with '" << name << "'\n";        \
   }(expr)
 
 // Static initialization of CUDA context for device ordinal 0.
 static auto InitializeCtx = [] {
+  CUDA_REPORT_IF_ERROR(cuInit(/*flags=*/0));
   CUdevice device;
   CUDA_REPORT_IF_ERROR(cuDeviceGet(&device, /*ordinal=*/0));
   CUcontext context;
   CUDA_REPORT_IF_ERROR(cuCtxCreate(&context, /*flags=*/0, device));
   return 0;
 }();
 
 extern "C" CUmodule mgpuModuleLoad(void *data) {

mlir/tools/mlir-rocm-runner/rocm-runtime-wrappers.cpp

[](hipError_t result) {                                                      \
     const char *name = hipGetErrorName(result);                                \
     if (!name)                                                                 \
       name = "<unknown>";                                                      \
     llvm::errs() << "'" << #expr << "' failed with '" << name << "'\n";        \
   }(expr)
 
 // Static initialization of HIP context for device ordinal 0.
 static auto InitializeCtx = [] {
+  HIP_REPORT_IF_ERROR(hipInit(/*flags=*/0));
   hipDevice_t device;
   HIP_REPORT_IF_ERROR(hipDeviceGet(&device, /*ordinal=*/0));
   hipContext_t context;
   HIP_REPORT_IF_ERROR(hipCtxCreate(&context, /*flags=*/0, device));
   return 0;
 }();
 
 extern "C" hipModule_t mgpuModuleLoad(void *data) {