Cut some needless muts in compiler

[nickdrozd]

Hello. This PR cuts most of the unnecessary &mut references from the compiler, changing them to plain & references instead. There are also a few resulting mut variable flags cut as well.

Should cause no change in behavior. Possibly will make something along the way a little faster?

x check and x clippy passed locally.

[rustbot]

r? @jackh726

rustbot has assigned @jackh726.
They will have a look at your PR within the next two weeks and either review your PR or reassign to another reviewer.

Use r? to explicitly pick a reviewer

[rustbot]

rustc_errors::annotate_snippet_emitter_writer was changed

cc @Muscraft

rustc_errors::emitter was changed

cc @Muscraft

Some changes occurred in compiler/rustc_builtin_macros/src/autodiff.rs

cc @ZuseZ4

Some changes occurred in compiler/rustc_codegen_ssa

cc @WaffleLapkin

Some changes occurred to MIR optimizations

cc @rust-lang/wg-mir-opt

Some changes occurred in match lowering

cc @Nadrieril

Some changes occurred to the CTFE / Miri interpreter

cc @rust-lang/miri

changes to the core type system

cc @compiler-errors, @lcnr

Some changes occurred in coverage instrumentation.

cc @Zalathar

Some changes occurred to the CTFE machinery

cc @RalfJung, @oli-obk, @lcnr

Some changes occurred in compiler/rustc_attr_parsing

cc @jdonszelmann

Some changes occurred in rustc_ty_utils::consts.rs

cc @BoxyUwU

Some changes occurred in need_type_info.rs

cc @lcnr

Some changes occurred in match checking

cc @Nadrieril

Some changes occurred to constck

cc @fee1-dead

[RalfJung]

Please keep the mut here, we don't know which functions will be hooked here in the future and they may need mutable access.

[lcnr]

While I think some of these changes are desirable and cleanup the code and it's helpful to state "this doesn't actually mutate the visitor", there are quite a few cases where whether they mutate some context is an implementation detail and the fact that they currently don't isn't actually meaningful and would result in churn.

e.g. in confirmation all confirm_X functions should all be able to mutate the SelectionContext and the fact that they don't right now isn't meaningful. If we change them in the future to do so we now need to know whether adding mut here is fine, readd it, and then hope that nobody used them in ways where that's annoying.

I don't think removing them based on a mechanical check for "is the mut currently required" is worth it. There are a few which seem like clear improvements, but a general blanket change feels ambiguous (and likely to cause merge conflicts) enough that I would prefer not merging this.