[Migrated] For loops need more optimizations

Issue automatically imported from old repo: https://github.com/EmbarkStudios/rust-gpu/issues/739
Old labels: t: enhancement,a: codegen
Originally creatd by hrydgard on 2021-08-30T12:50:21Z

Due to how for loops over a range work in Rust, a lot of scaffolding code is generated that is then expected to be optimized away, but in rust-gpu, that doesn't happen enough yet. We had a case where I had to change a for i in 0..8 {} to a while i < 8 { ...; i++ } loop to work around SPIRV-Cross making a right mess out of it.

Filing this issue to have something to point to from the code comment, mainly, not expectations for a quick resolution as I know it's a thorny one.

La issue cita i cicli for basati su intervalli di Rust, la generazione del codice di rust-gpu e SPIRV-Cross, ma non fornisce alcun file sorgente, test o punto di ingresso. Inizia individuando il commento nel codice a cui si fa riferimento e il percorso di codegen per i cicli basati su intervalli; il completamento richiederebbe di ridurre il codice ausiliario generato e confermare che SPIRV-Cross produca un output accettabile senza la soluzione alternativa con il ciclo while.