Since employing ccache on our CI server, we find that the bottleneck in terms of build time is now our static analysis pass, that uses clang-tidy, among other tools. Does anyone know of a way to accelerate clang-tidy in a similar way to how ccache does so with a regular compiler?