llama.cpp

439e68c1 - cmake : re-enable GCC -Wshadow