mirror of
https://github.com/rust-lang/rust.git
synced 2025-11-25 12:07:08 +00:00
This allows LLVM to optimize comparisons to zero before & after the multiplication into one, saving on code size and eliminating an (always true) branch from most Vec allocations.