Spawned off of https://github.com/rust-lang/rust/pull/96214#issuecomment-1104662390 Basically, I'm trying to understand how the PR sequence of rust-lang/rust#95379; rust-lang/rust#96020; rust-lang/rust#96214 ended up with both PR rust-lang/rust#96020 *and* PR rust-lang/rust#96214 being flagged as regressions. From my perspective, the data next to no change between rust-lang/rust#96020 and rust-lang/rust#96214 on this benchmark, at least from eyeballing the graph:  (but it could be that I'm very tired and overlooking something)