Laurențiu Nicola

Looking at GEOS, the [sweepline implementation](https://github.com/libgeos/geos/tree/main/src/index/sweepline) there seems surprisingly bare-bones. I can't find that comment any more, but a user suggested that we're running into this bug because we recompute...

Ouch: ``` d = 2 => 111 d = 9 => 113 d = 10 => 118 d = 11 => 109 d = 100 => 113 ``` So.. you...

> Switch from an LRU cache to a scan-resistant cache (LRU-K) Regarding this, I think the original Salsa used a scan-resistant strategy (SLRU?), see https://github.com/rust-lang/rust-analyzer/blob/ac389ce2ef9fa65edd3aa840648ae4a6637a4040/crates/salsa/src/lru.rs#L228, but it was dropped in...

Re flat syntax trees, leaving this here: https://github.com/saschagrunert/indextree.

There's nothing in https://github.com/rust-lang/rust-analyzer/pull/17880/files that could cause this.

@mihaiolteanu I assume you're not using rust-analyzer?

@djc humbly nominating this for 1.28.2.

Still fails in https://github.com/rust-lang/rust-analyzer/pull/19301. One difference might be the fact that the GHA runners already have an older rustup and Rust toolchain installed.

Another example from #17448: ```rust use itertools::izip; fn main() { let mut yp_sn = vec![0;5]; let yp = vec![0;5]; let x = vec![0;5]; for (x, yp_sn, _yp) in izip!(x, yp_sn.iter_mut(),...

Duplicate of https://github.com/rust-lang/rust-analyzer/issues/17739, but you can ignore the error.