There has been a discussion about CMake integration with SYCL on the CMake project but as yet it has not quite been able to progress but you might want to add your support to such an effort. https ://gitlab.kitware.com/cmake/cmake/-/issues/21711
In the meantime there are some useful files here https ://github.com/KhronosGroup/SYCL-CTS/tree/SYCL-1.2.1/master/cmake
If you are using DPC++, you may want to use
SET(CMAKE_CXX_FLAGS "-fsycl -fsycl-unnamed-lambda")
And this should be what you need.
However if you want to use DPC++ with the CUDA support you will need to use
SET(CMAKE_CXX_FLAGS "-fsycl -fsycl-targets=nvptx64-nvidia-cuda-sycldevice -fsycl-unnamed-lambda -Wno-unknown-cuda-version")
Alternatively you could use the script
and then launch CLion from that same shell so all the required environment variables are passed to CLion and it’s subprocesses, you just have to select the DPC++ compiler in your toolchain.