llama.cpp

b4e3dc61 - vulkan: add `v_dot2_f32_f16` support in matrix-matrix multiplication and Flash Attention (#24123)

vulkan: add `v_dot2_f32_f16` support in matrix-matrix multiplication and Flash Attention (#24123) * vulkan: add support for valve fp16 dot2 extension * use macro for dot2 path choice * properly check for the feature * add dot_product abstraction to reduce preprocessor branching