Signed-off-by: onur-ozkan <[email protected]>

We don't need to run `fn update_submodule` on rust-analyzer
as it's no longer a submodule.

Signed-off-by: onur-ozkan <[email protected]>

Fix rust-lang#101351.

When an associated type on a type parameter is used, and the type
parameter isn't constrained by the correct trait, suggest the
appropriate trait bound:

```
error[E0220]: associated type `Associated` not found for `T`
 --> file.rs:6:15
  |
6 |     field: T::Associated,
  |               ^^^^^^^^^^ there is a similarly named associated type `Associated` in the trait `Foo`
  |
help: consider restricting type parameter `T`
  |
5 | struct Generic<T: Foo> {
  |                 +++++
  ```

When an associated type on a type parameter has a typo, suggest fixing
it:

```
error[E0220]: associated type `Baa` not found for `T`
  --> $DIR/issue-55673.rs:9:8
   |
LL |     T::Baa: std::fmt::Debug,
   |        ^^^ there is a similarly named associated type `Bar` in the trait `Foo`
   |
help: change the associated type name to use `Bar` from `Foo`
   |
LL |     T::Bar: std::fmt::Debug,
   |        ~~~
```

This configuration will exclude rollup PRs and subtree sync PRs from
merge commit detection. On other PRs, it will post the default warning
message and add the `has-merge-commits` and `S-waiting-on-author`
labels when merge commits are detected.

The eventual vision is to have bors refuse to merge if the
`has-merge-commits` label is present. A reviewer can still
force the merge by removing that label if they so wish.

Enable triagebot no-merges check

Follow-up on rust-lang/triagebot#1704

### Motivation

Occasionally, a merge commit like rust-lang@cb5c011 makes it past manual review and gets merged into master.

At one point, we tried adding a check to CI to prevent this from happening (rust-lang#105058), but that ended up [problematic](rust-lang#106319 (comment)) and was [reverted](rust-lang#106320). This kind of check is simply too fragile for CI, and there must be a way for a human to override the bot's decision.

The capability to detect and warn about merge commits has been present in triagebot for quite some time, but was never enabled at rust-lang/rust, possibly due to concerns about false positives on rollup and subtree sync PRs. This PR intends to alleviate those concerns.

### Configuration

This configuration will exclude rollup PRs and subtree sync PRs from merge commit detection, and it will post the default warning message and add the `has-merge-commits` and `S-waiting-on-author` labels when merge commits are detected on other PRs.

The eventual vision is to have bors refuse to merge if the `has-merge-commits` label is present. A reviewer can still force the merge by removing that label if they so wish.

### Note for contributors

The rollup tool should add that label automatically, but anyone performing subtree updates should begin including "subtree update" in the titles of those PRs, to avoid false positives.

r? infra

## Open Questions

1. This configuration uses the default message that's built into triagebot:

> There are merge commits (commits with multiple parents) in your changes. We have a [no merge policy](https://rustc-dev-guide.rust-lang.org/git.html#no-merge-policy) so these commits will need to be removed for this pull request to be merged.
>
> You can start a rebase with the following commands:
> ```shell-session
> $ # rebase
> $ git rebase -i master
> $ # delete any merge commits in the editor that appears
> $ git push --force-with-lease
> ```

Any changes to this are easy, I'll just have to add a `message` option. Should we mention the excluded titles in the message?

Suggest trait bounds for used associated type on type param

Fix rust-lang#101351.

When an associated type on a type parameter is used, and the type parameter isn't constrained by the correct trait, suggest the appropriate trait bound:

```
error[E0220]: associated type `Associated` not found for `T`
 --> file.rs:6:15
  |
6 |     field: T::Associated,
  |               ^^^^^^^^^^ there is a similarly named associated type `Associated` in the trait `Foo`
  |
help: consider restricting type parameter `T`
  |
5 | struct Generic<T: Foo> {
  |                 +++++
  ```

When an associated type on a type parameter has a typo, suggest fixing
it:

```
error[E0220]: associated type `Baa` not found for `T`
  --> $DIR/issue-55673.rs:9:8
   |
LL |     T::Baa: std::fmt::Debug,
   |        ^^^ there is a similarly named associated type `Bar` in the trait `Foo`
   |
help: change the associated type name to use `Bar` from `Foo`
   |
LL |     T::Bar: std::fmt::Debug,
   |        ~~~
```

…urces, r=clubby789

vendoring in tarball sources

fixes rust-lang#94782

…notriddle

Update minifier version to 0.2.3

Thanks for the fix `@notriddle` !

r? `@notriddle`

Update my mailmap entry

…r=Kobzol

opt-dist: disable unused features for tabled crate

Features looks unused, so left only used ones.

r? `@Kobzol`

…r=oli-obk

Remove `DefiningAnchor::Bubble` from opaque wf check

Set the defining anchor to `DefiningAnchor::Bind(parent_def_id)` where `parent_def_id` is the first parent def-id that isn't an opaque.

This "fixes" some of the nested-return-type wf tests. If we *do* want these to be hard-errors for TAITs, we should probably make those error separately from this check (i.e. via some check like the code in the `OPAQUE_HIDDEN_INFERRED_BOUND` lint). The fact that some of these tests fail but not all of them seems kinda coincidental.

r? oli-obk