Remove fold code and add `Const::internal()` to StableMIR
We are not planning to support user generated constant in the foreseeable future, so we are cleaning up the fold logic and user created type for now. Users should use `Instance::resolve` in order to trigger monomorphization.
The Instance::resolve was however incomplete, since we weren't handling internalizing constants yet. Thus, I added that.
I decided to keep the `Const` fields private in case we decide to translate them lazily.
Modernize rustc_builtin_macros generics helpers
- Rustfmt-compatible formatting for the code snippets in comments
- Eliminate an _"Extra scope required"_ obsoleted by NLL
Uplift `ClauseKind` and `PredicateKind` into `rustc_type_ir`
Uplift `ClauseKind` and `PredicateKind` into `rustc_type_ir`.
Blocked on #116951
r? `@ghost`
Get rid of `'tcx` lifetime on `ConstVid`, `EffectVid`
These are simply newtyped numbers, so don't really have a reason (per se) to have a lifetime -- `TyVid` and `RegionVid` do not, for example.
The only consequence of this is that we need to use a new key type for `UnifyKey` that mentions `'tcx`. This is already done for `RegionVid`, with `RegionVidKey<'tcx>`, but this `UnifyKey` trait implementation may have been the original reason to give `ConstVid` a lifetime. See the changes to `compiler/rustc_middle/src/infer/unify_key.rs` specifically.
I consider the code cleaner this way, though -- we removed quite a few unnecessary `'tcx` in the process. This also makes it easier to uplift these two ids to `rustc_type_ir`, which I plan on doing in a follow-up PR.
r? `@BoxyUwU`
This requires introducing `tidy_error_ext!` as an alternative to
`tidy_error!`. It is passed a closure that is called for an error. This
lets tests capture tidy error messages in a buffer instead of printing
them to stderr.
It also requires pulling part of `check` out into a new function
`check_lines`, so that tests can pass in an iterator over some strings
instead of a file.
- Tweak some comments.
- No need to do the `concat!` trick on `START_MARKER` because it's
immediately followed by `END_MARKER`.
- Fix an off-by-one error in the line number for an error message.
- When a second start marker is found without an intervening end marker,
after giving an error, treat it as though it ends the section. It's
hard to know exactly what to do in this case, but it makes unit
testing this case a little simpler (in the next commit).
- If an end marker occurs without a preceding start marker, issue an
error.
We are not planning to support user generated constant in the
foreseeable future, so we are removing the Fold logic for now in
favor of the Instance::resolve logic.
The Instance::resolve was however incomplete, since we weren't handling
internalizing constants yet. Thus, I added that.
I decided to keep the Const fields private in case we decide to
translate them lazily.
These are comment lines in `Cargo.toml` files.
But exclude lines starting with `#!` from the skipping, because we want
to check them. (Rust `#![feature(...)]` lines.)
Also allow empty lines, which are occasionally useful.
Currently, if a `tidy-alphabetical-end` marker appears on the last line
of a file, tidy will erroneously issue a "reach end of file expecting
`tidy-alphabetical-end`" error. This is because it uses `take_while`
within `check_section`, which consumes the line with the end marker, and
then after `check_section` returns `check` peeks for at least one more
line, which won't be there is the marker was on the last line.
This commit fixes the problem, by removing the use of `take_while`, and
doing the "reached end of file" test within `check_section` without
using `peek`.
It also renames `{START,END}_COMMENT` as `{START,END}_MARKER`, which is
a more appropriate name.
