pub(crate) fn run_thin(
    cgcx: &CodegenContext<LlvmCodegenBackend>,
    modules: Vec<(String, ThinBuffer)>,
    cached_modules: Vec<(SerializedModule<ModuleBuffer>, WorkProduct)>
) -> Result<(Vec<LtoModuleCodegen<LlvmCodegenBackend>>, Vec<WorkProduct>), FatalError>
Expand description

Performs thin LTO by performing necessary global analysis and returning two lists, one of the modules that need optimization and another for modules that can simply be copied over from the incr. comp. cache.