Fail to compile DPC++ toolchain with support for NVIDIA CUDA

[hillyuan1]

I am tring to compile DPC++ from source code. After clone the source code from https://github.com/intel/llvm. The configuration works well follows

```

CUDA_LIB_PATH=/home/app/cuda/12.1/lib64/stubs CC=gcc CXX=g++ python ./llvm/buildbot/configure.py
--cuda --cmake-opt="-DCUDA_TOOLKIT_ROOT_DIR=/home/app/cuda/12.1"

```

but fails when compiled by gcc7.4.0

```

CUDA_LIB_PATH=/path/to/cuda/toolkit/lib64/stubs CC=gcc CXX=g++ python ./llvm/buildbot/compile.py

```

The error message says

```

/sycl_workspace/llvm/sycl/source/builtins/relational_functions.cpp: In function ‘int sycl::_V1::__isinf_impl(double)’:
/sycl_workspace/llvm/sycl/source/builtins/host_helper_macros.hpp:42:5: error: static assertion failed
static_assert(std::is_same_v<decltype(NAME##_host_impl(NUM_ARGS##_ARG)), \

```

Attached in my configure file.  I want to know how to solve this problem. Thanks for your help.

Best regards

X. YUAN, Ph.D in Solid Mechanics

[Alex_Y_Intel]

This is for your reference: 
https://github.com/intel/llvm/blob/sycl/sycl/doc/GetStartedGuide.md#build-dpc-toolchain-with-support-for-nvidia-cuda