[llama-cpp] add new port (and its ggml dependency)

[cenit]

• Changes comply with the maintainer guide.
• The name of the port matches an existing name for this component on https://repology.org/ if possible, and/or is strongly associated with that component on search engines.
• Optional dependencies are resolved in exactly one way. For example, if the component is built with CMake, all find_package calls are REQUIRED, are satisfied by vcpkg.json's declared dependencies, or disabled with CMAKE_DISABLE_FIND_PACKAGE_Xxx.
• The versioning scheme in vcpkg.json matches what upstream says.
• The license declaration in vcpkg.json matches what upstream says.
• The installed as the "copyright" file matches what upstream says.
• The source code of the component installed comes from an authoritative source.
• The generated "usage text" is accurate. See adding-usage for context.
• The version database is fixed by rerunning ./vcpkg x-add-version --all and committing the result.
• Only one version is in the new port's versions file.
• Only one version is added to each modified port's versions file.

[Neumann-A]

I had additionally:

vcpkg_cmake_config_fixup(CONFIG_PATH lib/cmake/Llama PACKAGE_NAME llama)
vcpkg_replace_string("${CURRENT_PACKAGES_DIR}/share/llama/LlamaConfig.cmake" "/../../../" "/../../")
vcpkg_replace_string("${CURRENT_PACKAGES_DIR}/share/llama/LlamaConfig.cmake" "${CURRENT_INSTALLED_DIR}" "\${PACKAGE_PREFIX_DIR}")

I think it has hardcoded paths to the CUDA libs somewhere?

[cenit]

@Neumann-A thanks for the suggestion. This is, in fact, just a quick scaffolding. Why didn't you push the port here? I was really sure to find it here already (by you or someone else, it's such a common lib nowadays 😄 )

[cenit]

in any case I will work on the port in the following days, with the goal of reaching green light everywhere, in case you don't have already your version working and you're good to go upstreaming it