Wild linker's benchmark gap with Mold traced to test setup, not speed
Benchmarking Wild vs. Mold
When Mold's updated linker benchmarks showed Wild substantially slower, Wild's author investigated. The main culprit: Mold's benchmarks delete the output file and use ext4, while Wild's used tmpfs and kept the file. Wild also lacked OS-specific tweaks like fallocate and hugepages, now added. After reproducing Mold's results on an M1, the author found Mold's recent 2.42.0/2.42.1 releases got considerably faster, explaining the rest of the gap.
Wild, at least the version benchmarked here, does best when allowed to fork and when the output already exists and is on tmpfs. i.e. the opposite of the configuration used in the Mold benchmarks.