Add internal lint against Ty == Ty

[Noratrieb]

== is often the first and most obvious operation people use to check whether types are equal, but it's usually not what they want, as it doesn't consider inference variables, bound vars and normalization.

We still need Ty: Eq for various things, for example for the query system.
There are still places where using it is "good enough" if the types are known to not have the caveats above (it should NOT be used in diagnostics though as the caveats will lead to worse diagnostics).

The lint points people to a yet-to-be-written section of the rustc dev guide about how to compare types properly.

I have not yet activated it by default because all the cfg_attr(bootstrap, allow) would be really annoying, so we should only activate and fix it after the bootstrap compiler is bumped to contain the lint.

[rustbot]

r? @eholk

(rustbot has picked a reviewer for you, use r? to override)

[jyn514]

I have not yet activated it by default because all the cfg_attr(bootstrap, allow) would be really annoying, so we should only activate and fix it after the bootstrap compiler is bumped to contain the lint.

Can you open an issue to enable it by default in 6 weeks?

[JakobDegen]

I have not yet activated it by default because all the cfg_attr(bootstrap, allow) would be really annoying

How many times does this fire on the repo? How many of those are things that should be fixed?

[Noratrieb]

I've done a little more thinking about this and looked at some code and I think there are now a few things to do:

• First, land this PR with the lint being disabled so I don't have to think about cfg(bootstrap)
• Then after the bootstrap bump, turn it on
• With the lint enabled, make rustc warning-free again. This is done through three ways.
  • In some crates like rustc_codegen_llvm, #![allow] it in the whole crate
  • For comparisons that are fine, add a new method intern_eq (bikeshed name later) that serves as "I know what I'm doing and want this"
  • For cases that seems sketchy, investigate them and fix them. This may take some time and can be nicely split up. Just #[allow] theses cases at first and add FIXMEs, these can be checked and then either replaced with intern_eq or some better comparison later

[compiler-errors]

r=me, one nit

[bors]

☔ The latest upstream changes (presumably #109001) made this pull request unmergeable. Please resolve the merge conflicts.

[anden3]

Hello @Nilstrieb! I noticed that this PR has some merge conflicts, what's the status of it?

[Noratrieb]

basically the steps described above. I need to fix the (absolutely trivial, thanks git) conflicts and then merge it and go through the plan later. I'll try doing that later this week if I don't forget about it.

[Noratrieb]

great question, i had to figure that out myself
I rebased and updated it to properly implement the plan described in #107393 (comment)
@rustbot ready (i think?)

[Noratrieb]

r? compiler-errors you've already reviewed it and we were talking about it

[compiler-errors]

r=me if ci is happy

[compiler-errors]

@rustbot author

[bors]

☔ The latest upstream changes (presumably #113734) made this pull request unmergeable. Please resolve the merge conflicts.

[rust-log-analyzer]

The job x86_64-gnu-tools failed! Check out the build log: (web) (plain)

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

GITHUB_ENV=/home/runner/work/_temp/_runner_file_commands/set_env_335d9cec-98a8-43ea-83a4-e218837fd9d6
GITHUB_EVENT_NAME=pull_request
GITHUB_EVENT_PATH=/home/runner/work/_temp/_github_workflow/event.json
GITHUB_GRAPHQL_URL=https://api.github.com/graphql
GITHUB_HEAD_REF=Are the types equalॽ Who even knows at this point
GITHUB_JOB=pr
GITHUB_PATH=/home/runner/work/_temp/_runner_file_commands/add_path_335d9cec-98a8-43ea-83a4-e218837fd9d6
GITHUB_REF=refs/pull/107393/merge
GITHUB_REF_NAME=107393/merge
GITHUB_REF_PROTECTED=false
---
   Compiling rustdoc v0.0.0 (/checkout/src/librustdoc)
error: using the a comparison operator on `Ty`
    --> src/librustdoc/clean/mod.rs:1358:32
     |
1358 |                 if self_arg_ty == self_ty {
     |
     |
     = note: this does probably not what you want as it does not handle inference variables and more
     = note: it's also not recommended to use it for diagnostics
     = help: for more information, see https://rustc-dev-guide.rust-lang.org/ty.html#comparing-types
     = note: `-D rustc::ty-compare-operator` implied by `-D warnings`
     = help: to override `-D warnings` add `#[allow(rustc::ty_compare_operator)]`
error: using the a comparison operator on `Ty`
    --> src/librustdoc/clean/mod.rs:1361:27
     |
1361 |                     if ty == self_ty {
1361 |                     if ty == self_ty {
     |                           ^^
     |
     = note: this does probably not what you want as it does not handle inference variables and more
     = note: it's also not recommended to use it for diagnostics
     = help: for more information, see https://rustc-dev-guide.rust-lang.org/ty.html#comparing-types
error: using the a comparison operator on `Ty`
   --> src/librustdoc/clean/simplify.rs:123:45
    |
    |
123 |                 if pred.trait_ref.self_ty() == self_ty { Some(pred.def_id()) } else { None }
    |
    |
    = note: this does probably not what you want as it does not handle inference variables and more
    = note: it's also not recommended to use it for diagnostics
    = help: for more information, see https://rustc-dev-guide.rust-lang.org/ty.html#comparing-types
error: using the a comparison operator on `Ty`
   --> src/librustdoc/passes/collect_intra_doc_links.rs:802:38
    |
    |
802 |             let saw_impl = impl_type == ty
    |
    |
    = note: this does probably not what you want as it does not handle inference variables and more
    = note: it's also not recommended to use it for diagnostics
    = help: for more information, see https://rustc-dev-guide.rust-lang.org/ty.html#comparing-types
error: could not compile `rustdoc` (lib) due to 4 previous errors
Build completed unsuccessfully in 0:10:09
cat: /tmp/toolstate/toolstates.json: No such file or directory
  local time: Sun Oct 15 15:17:30 UTC 2023

[bors]

☔ The latest upstream changes (presumably #116196) made this pull request unmergeable. Please resolve the merge conflicts.

[JohnCSimon]

Ping from triage
@Nilstrieb This PR is listed as approved but has a broken build.

[Noratrieb]

yes, i need to get back to this, there are some weird interactions with rustdoc that i haven't looked at

[JohnCSimon]

@Nilstrieb

Ping from triage: I'm closing this due to inactivity, Please reopen when you are ready to continue with this.
Note: if you are going to continue please open the PR BEFORE you push to it, else you won't be able to reopen - this is a quirk of github.
Thanks for your contribution.

@rustbot label: +S-inactive