[voldacar]

If I may ask, what was the use case for this code that they cared so much about a 2% difference in benchmarks? Aerospace? Game engine? Packet routing?

[sbierwagen]

I wouldn't expect aerospace, since I have been told embedded programmers in that field routinely disable compiler optimization, in the chance that a compiler bug or overzealous UB exploitation might introduce a bug into previously working code. Hard realtime requirements demand fast code, but not necessarily efficient code.