llama.cpp

7596487b - cmake : try to fix sycl+intel build (#9487)