Unreserve braced enum variants in value namespace

[petrochenkov]

With this PR braced enum variants (enum E { V { /*...*/ } }) no longer take a slot in value namespace, so the special case mentioned in the note in https://github.com/rust-lang/rfcs/blob/master/text/1506-adt-kinds.md#braced-structs is removed.

Report - #103578 (comment).

[rust-highfive]

r? @nagisa

(rust-highfive has picked a reviewer for you, use r? to override)

[rustbot]

Some changes occurred in src/librustdoc/clean/types.rs

cc @camelid

Some changes occurred in src/tools/clippy

cc @rust-lang/clippy

[petrochenkov]

Change description for the lang team.

Tuple structs/variants and unit structs/variants occupy slots in two namespaces, like described in RFC 1506 "Clarify the relationships between various kinds of structs and variants".

struct S; // type + value (constructor constant)
struct S(u8); // type + value (constructor function)
struct S { field: u8 } // only type

enum E {
    V, // type + value (constructor constant)
    V(u8), // type + value (constructor function)
    V { field: u8 }, // type + value (???)
}

In addition to that braced variants (but not structs) occupy an additional slot in value namespace.
The value it refers to is unusable, it produces an error when referred to from an expression or pattern (it can be reexported around though, otherwise use wouldn't be usable with braced variants at all).

This special case for enum variants was introduced in early 2016 in #30882 as future proofing.
During the last 6 years no specific plans for putting these names into use were proposed.
Also we now have editions to introduce these names in the future if such proposals appear.

So this PR removes these dummy names from value namespace thus making the language more consistent and simplifying the compiler.

[nagisa]

This generally looks okay, sans the diagnostics regressions, but those would be out of scope for me to vet and somebody working on diagnostics would need to review those for how fine is it to let them through in whatever state they are at the time this becomes ready for landing.

One thing worth doing now is a perf run though.

[petrochenkov]

the diagnostics regressions

I made some changes to keep the diagnostics more or less the same.
Keeping all the suggestions/labels require porting some logic from main resolution pass (rustc_resolve) to type-relative resolution (rustc_hir_typeck), which is a more invasive change not directly related to unreserving braced enum variants.

[petrochenkov]

One thing worth doing now is a perf run though.

@bors try @rust-timer queue

[rust-timer]

Awaiting bors try build completion.

@rustbot label: +S-waiting-on-perf

[bors]

⌛ Trying commit 3cd977e5816e4ce2a0ab1ee16ffda3000ccd4e45 with merge b0e9defe47ced648aa90c2ebf8a2fc7e64ea143a...

[bors]

☀️ Try build successful - checks-actions
Build commit: b0e9defe47ced648aa90c2ebf8a2fc7e64ea143a (b0e9defe47ced648aa90c2ebf8a2fc7e64ea143a)

[rust-log-analyzer]

A job failed! Check out the build log: (web) (plain)

Click to see the possible cause of the failure (guessed by this bot)

[bors]

💔 Test failed - checks-actions

[petrochenkov]

@bors retry

[bors]

⌛ Testing commit 7a5376d with merge b7463e8...

[bors]

☀️ Test successful - checks-actions
Approved by: nagisa
Pushing b7463e8 to master...

[rust-timer]

Finished benchmarking commit (b7463e8): comparison URL.

Overall result: ✅ improvements - no action needed

@rustbot label: -perf-regression

Instruction count

This is a highly reliable metric that was used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-0.4%	[-0.4%, -0.4%]	1
All ❌✅ (primary)	-	-	0

Max RSS (memory usage)

Results

This is a less reliable metric that may be of interest but was not used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-1.8%	[-1.8%, -1.8%]	1
All ❌✅ (primary)	-	-	0

Cycles

Results

This is a less reliable metric that may be of interest but was not used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-2.5%	[-2.5%, -2.5%]	1
All ❌✅ (primary)	-	-	0