Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Expose all OS-specific modules in libstd doc. #43348

Merged
merged 4 commits into from
Aug 13, 2017

Conversation

kennytm
Copy link
Member

@kennytm kennytm commented Jul 20, 2017

  1. Uses the special --cfg dox configuration passed by rustbuild when running rustdoc. Changes the #[cfg(platform)] into #[cfg(any(dox, platform))] so that platform-specific API are visible to rustdoc.

  2. Since platform-specific implementations often won't compile correctly on other platforms, rustdoc is changed to apply everybody_loops to the functions during documentation and doc-test harness.

  3. Since platform-specific code are documented on all platforms now, it could confuse users who found a useful API but is non-portable. Also, their examples will be doc-tested, so must be excluded when not testing on the native platform. An undocumented attribute #[doc(cfg(...))] is introduced to serve the above purposed.

Fixes #24658 (Does not fully implement #1998).

@rust-highfive
Copy link
Collaborator

r? @alexcrichton

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

@alexcrichton alexcrichton added the S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. label Jul 20, 2017
@carols10cents
Copy link
Member

friendly ping for review @alexcrichton! also pinging you on irc!

@alexcrichton
Copy link
Member

Ah sorry for the delay, but thanks for the PR @kennytm! This has defnitely long since been one of the major issues with the documentation of the standard library, and it's exciting to see movement towards fixing it!

This is a pretty big PR though is sort of difficult to review, and it's unclear to me at least how easy it'll be to maintain this into the future. I wonder if we could pursue a slightly smaller implementation in the meantime which may have practically the same impact? It looks like a lot of the complication here comes from handling the MetadataExt trait in various modules, but since those are all deprecated anyway I wonder if we could avoid them?

The "biggies" to handle here I think are std::os::{unix, windows} and I think the support needed for the may be a bit smaller? What do you think?

One other worry I'd have is that I think we'll want to do an audit of the documentation to ensure that everything is clearly documented as either Unix or Windows specific. Otherwise it may be confusing to see those windows methods in the Unix docs!

Also cc @rust-lang/libs, others may be interested in this as well!

@kennytm kennytm force-pushed the fix-24658-doc-every-platform branch from 0923acd to 5e89b23 Compare July 28, 2017 13:51
@kennytm
Copy link
Member Author

kennytm commented Jul 28, 2017

@alexcrichton Thanks for the comments :)

I've reverted most of the MetadataExt change. Now only std::os::linux is exposed, just like before. But I don't think MetadataExt the trait is deprecated; only the as_raw_stat() method is deprecated.

(Additionally, the documented_os_specific_impl! macro is downsized a lot thanks to the new stage0.)

This PR only labels OsStrExt/OsStringExt with the "Windows only" and "Unix only" labels, since I believe these will confuse most people. There could be a special badge for these non-portable function, like the stability badge.

Ideally this should be handled by rustdoc by checking the #[cfg], perhaps after #41619 (rust-lang/rfcs#1868 "a portability lint") is implemented. The RFC would also render this PR and the whole std::os module obsolete.

(↓ Not implemented in this PR)

@alexcrichton
Copy link
Member

Oh right yes, MetadataExt is indeed not deprecated!

Reading over this again though I'm still pretty wary personally. This seems very meticulously crafted in the sense that it's going to be pretty difficult to maintain over time. For example I could see that perhaps impls are added which involve adding more and more cases to the documented_os_specific_impl macro (or other related macros). It would be great to solve this problem though!

I wonder if perhaps we pursue solutions in rustdoc other than the standard library? For example there's no real need for rustdoc to actually typecheck these function bodies. Could rustdoc unconditionally always replace all function bodies with loop {}? That would remove the need for the macro I think, right?

I think we may want to also pursue a solution in rustdoc for the code examples here. For example we could have something like unix in the three backticks (or windows) to indicate that the example only runs on one platform maybe? Or better yet we could have something like #[doc(cfg(unix))] which could be an unstable attribute for now but indicate that rustdoc should (a) only compile examples on that platform and (b) warn in the docs that the function is unix-only in a uniform fashion.

Overall I'm just wondering if we could reduce the maintenance burden on the standard library as we add more impls/examples here over time!

@kennytm
Copy link
Member Author

kennytm commented Jul 31, 2017

@alexcrichton Sounds good to me! That could get rid of documented_os_specific_impl, but I think cfg(dox) still have the stay, right?

(cc @GuillaumeGomez @steveklabnik about the potential rustdoc changes)

@alexcrichton
Copy link
Member

Yeah we'll definitely still need #[cfg(dox)] in one way or another, I'm just hoping that we can reduce the amount of #[cfg(dox)] to O(1) basically in terms of some constant overhead but not needing annotation per-function (ideally at least)

@GuillaumeGomez
Copy link
Member

Wo, that's quite the improvement! Consider I'm all 👍

@kennytm kennytm changed the title Expose all OS-specific modules in libstd doc. [WIP] Expose all OS-specific modules in libstd doc. Aug 4, 2017
@kennytm kennytm force-pushed the fix-24658-doc-every-platform branch 2 times, most recently from c74278b to 768f6da Compare August 5, 2017 07:34
@kennytm kennytm changed the title [WIP] Expose all OS-specific modules in libstd doc. Expose all OS-specific modules in libstd doc. Aug 5, 2017
@kennytm
Copy link
Member Author

kennytm commented Aug 5, 2017

@alexcrichton Updated to include everybody_loops (in 622248a) and #[doc(cfg(...))] (in 3dea51f1eb).

I did not feature-gate doc(cfg(...)) since all other doc() attributes like doc(no_default_passes) are not.

Current output. cc @GuillaumeGomez some CSS changed.


1-fs8


2-fs8


Fullpage screenshot of `Metadata` structure

3-fullpage-fs8

@kennytm kennytm force-pushed the fix-24658-doc-every-platform branch from 768f6da to a59eb52 Compare August 5, 2017 08:08
@GuillaumeGomez
Copy link
Member

Seems good to me (for the CSS part).

@arielb1
Copy link
Contributor

arielb1 commented Aug 8, 2017

@alexcrichton - are you looking at reviewing this? Friendly ping to make sure this isn't getting lost.

@aturon
Copy link
Member

aturon commented Aug 8, 2017

This is awesome!

@alexcrichton
Copy link
Member

This also looks fantastic to me, thanks so much for taking on the refactoring @kennytm!

Only two thoughts now at this point:

  • Would it be easy to feature-gate the #[doc(cfg(..))] attribute for now? Just to not break code by accident if we change this in the future.
  • Could a few rustdoc tests be added for #[doc(cfg(..))] and what it renders? Namely the inheritance logic and such.

@QuietMisdreavus
Copy link
Member

I went ahead and checked out this PR to render an example std doc bundle with it: https://tonberry.quietmisdreavus.net/std-43348/std/os/index.html (This is rendered on a Linux server, in case that's useful information.)

@kennytm
Copy link
Member Author

kennytm commented Aug 10, 2017

@alexcrichton Done :). Feature gate tracked in #43781.

@bors
Copy link
Contributor

bors commented Aug 10, 2017

☔ The latest upstream changes (presumably #43730) made this pull request unmergeable. Please resolve the merge conflicts.

This prevents compilation failure we want to document a platform-specific
module. Every function is replaced by `loop {}` using the same construct
as `--unpretty everybody_loops`.

Note also a workaround to rust-lang#43636 is included: `const fn` will retain their
bodies, since the standard library has quite a number of them.
This attribute has two effects:

1. Items with this attribute and their children will have the "This is
   supported on **** only" message attached in the documentation.

2. The items' doc tests will be skipped if the configuration does not
   match.
@kennytm kennytm force-pushed the fix-24658-doc-every-platform branch from 7228c96 to b4114eb Compare August 10, 2017 05:44
Manishearth added a commit to Manishearth/rust that referenced this pull request Jul 16, 2020
Don't run `everybody_loops` for rustdoc; instead ignore resolution errors

r? @eddyb
cc @petrochenkov, @GuillaumeGomez, @Manishearth, @ecstatic-morse, @marmeladema

~~Blocked on rust-lang#73743 Merged.
~~Blocked on crater run.~~ Crater popped up some ICEs ([now fixed](rust-lang#73566 (comment))). See [crater run](https://crater-reports.s3.amazonaws.com/pr-73566/index.html), [ICEs](rust-lang#73566 (comment)).
~~Blocked on rust-lang#74070 so that we don't make typeck_tables_of public when it shouldn't be.~~ Merged.

Closes rust-lang#71820, closes rust-lang#71104, closes rust-lang#65863.

## What is the motivation for this change?

As seen from a lengthy trail of PRs and issues (rust-lang#73532, rust-lang#73103, rust-lang#71820, rust-lang#71104), `everybody_loops` is causing bugs in rustdoc. The main issue is that it does not preserve the validity of the `DefId` tree, meaning that operations on DefIds may unexpectedly fail when called later. This is blocking intra-doc links (see rust-lang#73101).

This PR starts by removing `everybody_loops`, fixing rust-lang#71104 and rust-lang#71820. However, that brings back the bugs seen originally in rust-lang#43348: Since libstd documents items for all platforms, the function bodies sometimes do not type check. Here are the errors from documenting `libstd` with `everybody_loops` disabled and no other changes:

```rust
error[E0433]: failed to resolve: could not find `handle` in `sys`
  --> src/libstd/sys/windows/ext/process.rs:13:27
   |
13 |         let handle = sys::handle::Handle::new(handle as *mut _);
   |                           ^^^^^^ could not find `handle` in `sys`

error[E0425]: cannot find function `symlink_inner` in module `sys::fs`
   --> src/libstd/sys/windows/ext/fs.rs:544:14
    |
544 |     sys::fs::symlink_inner(src.as_ref(), dst.as_ref(), false)
    |              ^^^^^^^^^^^^^ not found in `sys::fs`

error[E0425]: cannot find function `symlink_inner` in module `sys::fs`
   --> src/libstd/sys/windows/ext/fs.rs:564:14
    |
564 |     sys::fs::symlink_inner(src.as_ref(), dst.as_ref(), true)
    |              ^^^^^^^^^^^^^ not found in `sys::fs`
```

## Why does this need changes to `rustc_resolve`?

Normally, this could be avoided by simply not calling the `typeck_item_bodies` pass. However, the errors above happen before type checking, in name resolution itself. Since name resolution is intermingled with macro expansion, and rustdoc needs expansion to happen before it knows all items to be documented, there needs to be someway to ignore _resolution_ errors in function bodies.

An alternative solution suggested by @petrochenkov was to not run `everybody_loops` on anything containing a nested `DefId`. This would solve some of the immediate issues, but isn't bullet-proof: the following functions still could not be documented if the items in the body failed to resolve:

- Functions containing a nested `DefId` (rust-lang#71104)
- ~~Functions returning `impl Trait` (rust-lang#43878 These ended up not resolving anyway with this PR.
- ~~`const fn`, because `loop {}` in `const fn` is unstable (rust-lang#43636 `const_loop` was just stabilized.

This also isn't exactly what rustdoc wants, which is to avoid looking at function bodies in the first place.

## What changes were made?

The hack implemented in this PR is to add an option to ignore all resolution errors in function bodies. This is enabled only for rustdoc. Since resolution errors are ignored, the MIR generated will be invalid, as can be seen in the following ICE:

```rust
error: internal compiler error: broken MIR in DefId(0:11 ~ doc_cfg[8787]::uses_target_feature[0]) ("return type"): bad type [type error]
  --> /home/joshua/src/rust/src/test/rustdoc/doc-cfg.rs:51:1
   |
51 | / pub unsafe fn uses_target_feature() {
52 | |     content::should::be::irrelevant();
53 | | }
   | |_^
```

Fortunately, rustdoc does not need to access MIR in order to generate documentation. Therefore this also removes the call to `analyze()` in `rustdoc::run_core`. This has the side effect of not generating all lints by default. Most lints are safe to ignore (does rustdoc really need to run liveness analysis?) but `missing_docs` in particular is disabled when it should not be. Re-running `missing_docs` specifically does not help, because it causes the typechecking pass to be run, bringing back the errors from rust-lang#24658:

```
error[E0599]: no method named `into_handle` found for struct `sys::unix::pipe::AnonPipe` in the current scope
  --> src/libstd/sys/windows/ext/process.rs:71:27
   |
71 |         self.into_inner().into_handle().into_raw() as *mut _
   |                           ^^^^^^^^^^^ method not found in `sys::unix::pipe::AnonPipe`
   |
```

Because of rust-lang#73743, we only run typeck on demand. So this only causes an issue for functions returning `impl Trait`, which were already special cased by `ReplaceFunctionWithBody`. However, it now considers `async fn f() -> T` to be considered `impl Future<Output = T>`, where before it was considered to have a concrete `T` type.

## How will this affect future changes to rustdoc?

- Any new changes to rustdoc will not be able to perform type checking without bringing back resolution errors in function bodies.
    + As a corollary, any new lints cannot require or perform type checking. In some cases this may require refactoring other parts of the compiler to perform type-checking only on-demand, see for example rust-lang#73743.
    + As a corollary, rustdoc can never again call `tcx.analysis()` unless this PR is reverted altogether.

## Current status

- ~~I am not yet sure how to bring back `missing_docs` without running typeck. @eddyb suggested allowing lints to opt-out of type-checking, which would probably be another rabbit hole.~~ The opt-out was implemented in rust-lang#73743. However, of the rustc lints, now _only_ missing_docs is run and no other lints: rust-lang#73566 (comment). We need a team decision on whether that's an acceptable tradeoff. Note that all rustdoc lints are still run (`intra_doc_link_resolution_failure`, etc). **UPDATE**: This was deemed acceptable in rust-lang#73566 (comment)
- ~~The implementation of optional errors in `rustc_resolve` is very brute force, it should probably be moved from `LateResolver` to `Resolver` to avoid duplicating the logic in many places.~~ I'm mostly happy with it now.

- This no longer allows errors in `async fn f() -> T`. This caused breakage in 50 crates out of a full crater run, all of which (that I looked at) didn't compile when run with rustc directly. In other words, it used to be that they could not be compiled but could still be documented; now they can't be documented either. This needs a decision from the rustdoc team on whether this is acceptable breakage. **UPDATE**: This was deemed acceptable in rust-lang#73566 (comment)
- ~~This makes `fn typeck_tables_of` in `rustc_typeck` public. This is not desired behavior, but needs the changes from rust-lang#74070 in order to be fixed.~~ Reverted.
Manishearth added a commit to Manishearth/rust that referenced this pull request Jul 16, 2020
Don't run `everybody_loops` for rustdoc; instead ignore resolution errors

r? @eddyb
cc @petrochenkov, @GuillaumeGomez, @Manishearth, @ecstatic-morse, @marmeladema

~~Blocked on rust-lang#73743 Merged.
~~Blocked on crater run.~~ Crater popped up some ICEs ([now fixed](rust-lang#73566 (comment))). See [crater run](https://crater-reports.s3.amazonaws.com/pr-73566/index.html), [ICEs](rust-lang#73566 (comment)).
~~Blocked on rust-lang#74070 so that we don't make typeck_tables_of public when it shouldn't be.~~ Merged.

Closes rust-lang#71820, closes rust-lang#71104, closes rust-lang#65863.

## What is the motivation for this change?

As seen from a lengthy trail of PRs and issues (rust-lang#73532, rust-lang#73103, rust-lang#71820, rust-lang#71104), `everybody_loops` is causing bugs in rustdoc. The main issue is that it does not preserve the validity of the `DefId` tree, meaning that operations on DefIds may unexpectedly fail when called later. This is blocking intra-doc links (see rust-lang#73101).

This PR starts by removing `everybody_loops`, fixing rust-lang#71104 and rust-lang#71820. However, that brings back the bugs seen originally in rust-lang#43348: Since libstd documents items for all platforms, the function bodies sometimes do not type check. Here are the errors from documenting `libstd` with `everybody_loops` disabled and no other changes:

```rust
error[E0433]: failed to resolve: could not find `handle` in `sys`
  --> src/libstd/sys/windows/ext/process.rs:13:27
   |
13 |         let handle = sys::handle::Handle::new(handle as *mut _);
   |                           ^^^^^^ could not find `handle` in `sys`

error[E0425]: cannot find function `symlink_inner` in module `sys::fs`
   --> src/libstd/sys/windows/ext/fs.rs:544:14
    |
544 |     sys::fs::symlink_inner(src.as_ref(), dst.as_ref(), false)
    |              ^^^^^^^^^^^^^ not found in `sys::fs`

error[E0425]: cannot find function `symlink_inner` in module `sys::fs`
   --> src/libstd/sys/windows/ext/fs.rs:564:14
    |
564 |     sys::fs::symlink_inner(src.as_ref(), dst.as_ref(), true)
    |              ^^^^^^^^^^^^^ not found in `sys::fs`
```

## Why does this need changes to `rustc_resolve`?

Normally, this could be avoided by simply not calling the `typeck_item_bodies` pass. However, the errors above happen before type checking, in name resolution itself. Since name resolution is intermingled with macro expansion, and rustdoc needs expansion to happen before it knows all items to be documented, there needs to be someway to ignore _resolution_ errors in function bodies.

An alternative solution suggested by @petrochenkov was to not run `everybody_loops` on anything containing a nested `DefId`. This would solve some of the immediate issues, but isn't bullet-proof: the following functions still could not be documented if the items in the body failed to resolve:

- Functions containing a nested `DefId` (rust-lang#71104)
- ~~Functions returning `impl Trait` (rust-lang#43878 These ended up not resolving anyway with this PR.
- ~~`const fn`, because `loop {}` in `const fn` is unstable (rust-lang#43636 `const_loop` was just stabilized.

This also isn't exactly what rustdoc wants, which is to avoid looking at function bodies in the first place.

## What changes were made?

The hack implemented in this PR is to add an option to ignore all resolution errors in function bodies. This is enabled only for rustdoc. Since resolution errors are ignored, the MIR generated will be invalid, as can be seen in the following ICE:

```rust
error: internal compiler error: broken MIR in DefId(0:11 ~ doc_cfg[8787]::uses_target_feature[0]) ("return type"): bad type [type error]
  --> /home/joshua/src/rust/src/test/rustdoc/doc-cfg.rs:51:1
   |
51 | / pub unsafe fn uses_target_feature() {
52 | |     content::should::be::irrelevant();
53 | | }
   | |_^
```

Fortunately, rustdoc does not need to access MIR in order to generate documentation. Therefore this also removes the call to `analyze()` in `rustdoc::run_core`. This has the side effect of not generating all lints by default. Most lints are safe to ignore (does rustdoc really need to run liveness analysis?) but `missing_docs` in particular is disabled when it should not be. Re-running `missing_docs` specifically does not help, because it causes the typechecking pass to be run, bringing back the errors from rust-lang#24658:

```
error[E0599]: no method named `into_handle` found for struct `sys::unix::pipe::AnonPipe` in the current scope
  --> src/libstd/sys/windows/ext/process.rs:71:27
   |
71 |         self.into_inner().into_handle().into_raw() as *mut _
   |                           ^^^^^^^^^^^ method not found in `sys::unix::pipe::AnonPipe`
   |
```

Because of rust-lang#73743, we only run typeck on demand. So this only causes an issue for functions returning `impl Trait`, which were already special cased by `ReplaceFunctionWithBody`. However, it now considers `async fn f() -> T` to be considered `impl Future<Output = T>`, where before it was considered to have a concrete `T` type.

## How will this affect future changes to rustdoc?

- Any new changes to rustdoc will not be able to perform type checking without bringing back resolution errors in function bodies.
    + As a corollary, any new lints cannot require or perform type checking. In some cases this may require refactoring other parts of the compiler to perform type-checking only on-demand, see for example rust-lang#73743.
    + As a corollary, rustdoc can never again call `tcx.analysis()` unless this PR is reverted altogether.

## Current status

- ~~I am not yet sure how to bring back `missing_docs` without running typeck. @eddyb suggested allowing lints to opt-out of type-checking, which would probably be another rabbit hole.~~ The opt-out was implemented in rust-lang#73743. However, of the rustc lints, now _only_ missing_docs is run and no other lints: rust-lang#73566 (comment). We need a team decision on whether that's an acceptable tradeoff. Note that all rustdoc lints are still run (`intra_doc_link_resolution_failure`, etc). **UPDATE**: This was deemed acceptable in rust-lang#73566 (comment)
- ~~The implementation of optional errors in `rustc_resolve` is very brute force, it should probably be moved from `LateResolver` to `Resolver` to avoid duplicating the logic in many places.~~ I'm mostly happy with it now.

- This no longer allows errors in `async fn f() -> T`. This caused breakage in 50 crates out of a full crater run, all of which (that I looked at) didn't compile when run with rustc directly. In other words, it used to be that they could not be compiled but could still be documented; now they can't be documented either. This needs a decision from the rustdoc team on whether this is acceptable breakage. **UPDATE**: This was deemed acceptable in rust-lang#73566 (comment)
- ~~This makes `fn typeck_tables_of` in `rustc_typeck` public. This is not desired behavior, but needs the changes from rust-lang#74070 in order to be fixed.~~ Reverted.
Manishearth added a commit to Manishearth/rust that referenced this pull request Jul 16, 2020
Don't run `everybody_loops` for rustdoc; instead ignore resolution errors

r? @eddyb
cc @petrochenkov, @GuillaumeGomez, @Manishearth, @ecstatic-morse, @marmeladema

~~Blocked on rust-lang#73743 Merged.
~~Blocked on crater run.~~ Crater popped up some ICEs ([now fixed](rust-lang#73566 (comment))). See [crater run](https://crater-reports.s3.amazonaws.com/pr-73566/index.html), [ICEs](rust-lang#73566 (comment)).
~~Blocked on rust-lang#74070 so that we don't make typeck_tables_of public when it shouldn't be.~~ Merged.

Closes rust-lang#71820, closes rust-lang#71104, closes rust-lang#65863.

## What is the motivation for this change?

As seen from a lengthy trail of PRs and issues (rust-lang#73532, rust-lang#73103, rust-lang#71820, rust-lang#71104), `everybody_loops` is causing bugs in rustdoc. The main issue is that it does not preserve the validity of the `DefId` tree, meaning that operations on DefIds may unexpectedly fail when called later. This is blocking intra-doc links (see rust-lang#73101).

This PR starts by removing `everybody_loops`, fixing rust-lang#71104 and rust-lang#71820. However, that brings back the bugs seen originally in rust-lang#43348: Since libstd documents items for all platforms, the function bodies sometimes do not type check. Here are the errors from documenting `libstd` with `everybody_loops` disabled and no other changes:

```rust
error[E0433]: failed to resolve: could not find `handle` in `sys`
  --> src/libstd/sys/windows/ext/process.rs:13:27
   |
13 |         let handle = sys::handle::Handle::new(handle as *mut _);
   |                           ^^^^^^ could not find `handle` in `sys`

error[E0425]: cannot find function `symlink_inner` in module `sys::fs`
   --> src/libstd/sys/windows/ext/fs.rs:544:14
    |
544 |     sys::fs::symlink_inner(src.as_ref(), dst.as_ref(), false)
    |              ^^^^^^^^^^^^^ not found in `sys::fs`

error[E0425]: cannot find function `symlink_inner` in module `sys::fs`
   --> src/libstd/sys/windows/ext/fs.rs:564:14
    |
564 |     sys::fs::symlink_inner(src.as_ref(), dst.as_ref(), true)
    |              ^^^^^^^^^^^^^ not found in `sys::fs`
```

## Why does this need changes to `rustc_resolve`?

Normally, this could be avoided by simply not calling the `typeck_item_bodies` pass. However, the errors above happen before type checking, in name resolution itself. Since name resolution is intermingled with macro expansion, and rustdoc needs expansion to happen before it knows all items to be documented, there needs to be someway to ignore _resolution_ errors in function bodies.

An alternative solution suggested by @petrochenkov was to not run `everybody_loops` on anything containing a nested `DefId`. This would solve some of the immediate issues, but isn't bullet-proof: the following functions still could not be documented if the items in the body failed to resolve:

- Functions containing a nested `DefId` (rust-lang#71104)
- ~~Functions returning `impl Trait` (rust-lang#43878 These ended up not resolving anyway with this PR.
- ~~`const fn`, because `loop {}` in `const fn` is unstable (rust-lang#43636 `const_loop` was just stabilized.

This also isn't exactly what rustdoc wants, which is to avoid looking at function bodies in the first place.

## What changes were made?

The hack implemented in this PR is to add an option to ignore all resolution errors in function bodies. This is enabled only for rustdoc. Since resolution errors are ignored, the MIR generated will be invalid, as can be seen in the following ICE:

```rust
error: internal compiler error: broken MIR in DefId(0:11 ~ doc_cfg[8787]::uses_target_feature[0]) ("return type"): bad type [type error]
  --> /home/joshua/src/rust/src/test/rustdoc/doc-cfg.rs:51:1
   |
51 | / pub unsafe fn uses_target_feature() {
52 | |     content::should::be::irrelevant();
53 | | }
   | |_^
```

Fortunately, rustdoc does not need to access MIR in order to generate documentation. Therefore this also removes the call to `analyze()` in `rustdoc::run_core`. This has the side effect of not generating all lints by default. Most lints are safe to ignore (does rustdoc really need to run liveness analysis?) but `missing_docs` in particular is disabled when it should not be. Re-running `missing_docs` specifically does not help, because it causes the typechecking pass to be run, bringing back the errors from rust-lang#24658:

```
error[E0599]: no method named `into_handle` found for struct `sys::unix::pipe::AnonPipe` in the current scope
  --> src/libstd/sys/windows/ext/process.rs:71:27
   |
71 |         self.into_inner().into_handle().into_raw() as *mut _
   |                           ^^^^^^^^^^^ method not found in `sys::unix::pipe::AnonPipe`
   |
```

Because of rust-lang#73743, we only run typeck on demand. So this only causes an issue for functions returning `impl Trait`, which were already special cased by `ReplaceFunctionWithBody`. However, it now considers `async fn f() -> T` to be considered `impl Future<Output = T>`, where before it was considered to have a concrete `T` type.

## How will this affect future changes to rustdoc?

- Any new changes to rustdoc will not be able to perform type checking without bringing back resolution errors in function bodies.
    + As a corollary, any new lints cannot require or perform type checking. In some cases this may require refactoring other parts of the compiler to perform type-checking only on-demand, see for example rust-lang#73743.
    + As a corollary, rustdoc can never again call `tcx.analysis()` unless this PR is reverted altogether.

## Current status

- ~~I am not yet sure how to bring back `missing_docs` without running typeck. @eddyb suggested allowing lints to opt-out of type-checking, which would probably be another rabbit hole.~~ The opt-out was implemented in rust-lang#73743. However, of the rustc lints, now _only_ missing_docs is run and no other lints: rust-lang#73566 (comment). We need a team decision on whether that's an acceptable tradeoff. Note that all rustdoc lints are still run (`intra_doc_link_resolution_failure`, etc). **UPDATE**: This was deemed acceptable in rust-lang#73566 (comment)
- ~~The implementation of optional errors in `rustc_resolve` is very brute force, it should probably be moved from `LateResolver` to `Resolver` to avoid duplicating the logic in many places.~~ I'm mostly happy with it now.

- This no longer allows errors in `async fn f() -> T`. This caused breakage in 50 crates out of a full crater run, all of which (that I looked at) didn't compile when run with rustc directly. In other words, it used to be that they could not be compiled but could still be documented; now they can't be documented either. This needs a decision from the rustdoc team on whether this is acceptable breakage. **UPDATE**: This was deemed acceptable in rust-lang#73566 (comment)
- ~~This makes `fn typeck_tables_of` in `rustc_typeck` public. This is not desired behavior, but needs the changes from rust-lang#74070 in order to be fixed.~~ Reverted.
Manishearth added a commit to Manishearth/rust that referenced this pull request Jul 16, 2020
Don't run `everybody_loops` for rustdoc; instead ignore resolution errors

r? @eddyb
cc @petrochenkov, @GuillaumeGomez, @Manishearth, @ecstatic-morse, @marmeladema

~~Blocked on rust-lang#73743 Merged.
~~Blocked on crater run.~~ Crater popped up some ICEs ([now fixed](rust-lang#73566 (comment))). See [crater run](https://crater-reports.s3.amazonaws.com/pr-73566/index.html), [ICEs](rust-lang#73566 (comment)).
~~Blocked on rust-lang#74070 so that we don't make typeck_tables_of public when it shouldn't be.~~ Merged.

Closes rust-lang#71820, closes rust-lang#71104, closes rust-lang#65863.

## What is the motivation for this change?

As seen from a lengthy trail of PRs and issues (rust-lang#73532, rust-lang#73103, rust-lang#71820, rust-lang#71104), `everybody_loops` is causing bugs in rustdoc. The main issue is that it does not preserve the validity of the `DefId` tree, meaning that operations on DefIds may unexpectedly fail when called later. This is blocking intra-doc links (see rust-lang#73101).

This PR starts by removing `everybody_loops`, fixing rust-lang#71104 and rust-lang#71820. However, that brings back the bugs seen originally in rust-lang#43348: Since libstd documents items for all platforms, the function bodies sometimes do not type check. Here are the errors from documenting `libstd` with `everybody_loops` disabled and no other changes:

```rust
error[E0433]: failed to resolve: could not find `handle` in `sys`
  --> src/libstd/sys/windows/ext/process.rs:13:27
   |
13 |         let handle = sys::handle::Handle::new(handle as *mut _);
   |                           ^^^^^^ could not find `handle` in `sys`

error[E0425]: cannot find function `symlink_inner` in module `sys::fs`
   --> src/libstd/sys/windows/ext/fs.rs:544:14
    |
544 |     sys::fs::symlink_inner(src.as_ref(), dst.as_ref(), false)
    |              ^^^^^^^^^^^^^ not found in `sys::fs`

error[E0425]: cannot find function `symlink_inner` in module `sys::fs`
   --> src/libstd/sys/windows/ext/fs.rs:564:14
    |
564 |     sys::fs::symlink_inner(src.as_ref(), dst.as_ref(), true)
    |              ^^^^^^^^^^^^^ not found in `sys::fs`
```

## Why does this need changes to `rustc_resolve`?

Normally, this could be avoided by simply not calling the `typeck_item_bodies` pass. However, the errors above happen before type checking, in name resolution itself. Since name resolution is intermingled with macro expansion, and rustdoc needs expansion to happen before it knows all items to be documented, there needs to be someway to ignore _resolution_ errors in function bodies.

An alternative solution suggested by @petrochenkov was to not run `everybody_loops` on anything containing a nested `DefId`. This would solve some of the immediate issues, but isn't bullet-proof: the following functions still could not be documented if the items in the body failed to resolve:

- Functions containing a nested `DefId` (rust-lang#71104)
- ~~Functions returning `impl Trait` (rust-lang#43878 These ended up not resolving anyway with this PR.
- ~~`const fn`, because `loop {}` in `const fn` is unstable (rust-lang#43636 `const_loop` was just stabilized.

This also isn't exactly what rustdoc wants, which is to avoid looking at function bodies in the first place.

## What changes were made?

The hack implemented in this PR is to add an option to ignore all resolution errors in function bodies. This is enabled only for rustdoc. Since resolution errors are ignored, the MIR generated will be invalid, as can be seen in the following ICE:

```rust
error: internal compiler error: broken MIR in DefId(0:11 ~ doc_cfg[8787]::uses_target_feature[0]) ("return type"): bad type [type error]
  --> /home/joshua/src/rust/src/test/rustdoc/doc-cfg.rs:51:1
   |
51 | / pub unsafe fn uses_target_feature() {
52 | |     content::should::be::irrelevant();
53 | | }
   | |_^
```

Fortunately, rustdoc does not need to access MIR in order to generate documentation. Therefore this also removes the call to `analyze()` in `rustdoc::run_core`. This has the side effect of not generating all lints by default. Most lints are safe to ignore (does rustdoc really need to run liveness analysis?) but `missing_docs` in particular is disabled when it should not be. Re-running `missing_docs` specifically does not help, because it causes the typechecking pass to be run, bringing back the errors from rust-lang#24658:

```
error[E0599]: no method named `into_handle` found for struct `sys::unix::pipe::AnonPipe` in the current scope
  --> src/libstd/sys/windows/ext/process.rs:71:27
   |
71 |         self.into_inner().into_handle().into_raw() as *mut _
   |                           ^^^^^^^^^^^ method not found in `sys::unix::pipe::AnonPipe`
   |
```

Because of rust-lang#73743, we only run typeck on demand. So this only causes an issue for functions returning `impl Trait`, which were already special cased by `ReplaceFunctionWithBody`. However, it now considers `async fn f() -> T` to be considered `impl Future<Output = T>`, where before it was considered to have a concrete `T` type.

## How will this affect future changes to rustdoc?

- Any new changes to rustdoc will not be able to perform type checking without bringing back resolution errors in function bodies.
    + As a corollary, any new lints cannot require or perform type checking. In some cases this may require refactoring other parts of the compiler to perform type-checking only on-demand, see for example rust-lang#73743.
    + As a corollary, rustdoc can never again call `tcx.analysis()` unless this PR is reverted altogether.

## Current status

- ~~I am not yet sure how to bring back `missing_docs` without running typeck. @eddyb suggested allowing lints to opt-out of type-checking, which would probably be another rabbit hole.~~ The opt-out was implemented in rust-lang#73743. However, of the rustc lints, now _only_ missing_docs is run and no other lints: rust-lang#73566 (comment). We need a team decision on whether that's an acceptable tradeoff. Note that all rustdoc lints are still run (`intra_doc_link_resolution_failure`, etc). **UPDATE**: This was deemed acceptable in rust-lang#73566 (comment)
- ~~The implementation of optional errors in `rustc_resolve` is very brute force, it should probably be moved from `LateResolver` to `Resolver` to avoid duplicating the logic in many places.~~ I'm mostly happy with it now.

- This no longer allows errors in `async fn f() -> T`. This caused breakage in 50 crates out of a full crater run, all of which (that I looked at) didn't compile when run with rustc directly. In other words, it used to be that they could not be compiled but could still be documented; now they can't be documented either. This needs a decision from the rustdoc team on whether this is acceptable breakage. **UPDATE**: This was deemed acceptable in rust-lang#73566 (comment)
- ~~This makes `fn typeck_tables_of` in `rustc_typeck` public. This is not desired behavior, but needs the changes from rust-lang#74070 in order to be fixed.~~ Reverted.
@jyn514 jyn514 added A-cross Area: Cross compilation F-doc_cfg `#![feature(doc_cfg)]` A-docs Area: documentation for any part of the project, including the compiler, standard library, and tools T-libs-api Relevant to the library API team, which will review and decide on the PR/issue. T-rustdoc Relevant to the rustdoc team, which will review and decide on the PR/issue. merged-by-bors This PR was explicitly merged by bors. and removed S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. labels Nov 21, 2020
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
A-cross Area: Cross compilation A-docs Area: documentation for any part of the project, including the compiler, standard library, and tools F-doc_cfg `#![feature(doc_cfg)]` merged-by-bors This PR was explicitly merged by bors. T-libs-api Relevant to the library API team, which will review and decide on the PR/issue. T-rustdoc Relevant to the rustdoc team, which will review and decide on the PR/issue.
Projects
None yet
Development

Successfully merging this pull request may close these issues.

std::os::windows missing on http://doc.rust-lang.org/nightly/std/os/