llama.cpp

f50db6ae - Add check for VK_KHR_portability_enumeration for MoltenVK support