rustdoc: Cleanup various `clean` types #88379

camelid · 2021-08-27T02:44:47Z

Cleanup various clean types.

jyn514

Got about halfway through and ran out of time - this is amazing work ❤️ not sure why it's marked as a draft, but feel free to ignore the comments until you're happy with it :)

src/librustdoc/html/render/mod.rs

src/librustdoc/clean/types.rs

jyn514 · 2021-08-27T13:48:03Z

src/librustdoc/clean/mod.rs

            res: self.res,
-            segments: if self.is_global() { &self.segments[1..] } else { &self.segments }.clean(cx),
+            segments: self.segments.clean(cx),


Hmm, what will segments[0] be for a global path? Just an empty string?

I thought kw::PathRoot, but it may not be because otherwise I'd expect tests to be failing. I'll have to look into it.

It will be kw::PathRoot, which renders as {{root}}, but I haven't found a way to make rustdoc (erroneously) print out {{root}}::foo::bar. (It may be possible though.)

src/librustdoc/clean/auto_trait.rs

jyn514 · 2021-08-27T14:00:18Z

src/librustdoc/html/format.rs

            // Paths like `T::Output` and `Self::Output` should be rendered with all segments.
-            resolved_path(f, did, path, is_generic, use_absolute, cx)
+            resolved_path(f, did, path, path.is_generic(), use_absolute, cx)


You don't need to pass in path.is_generic, you can calculate it in resolved_path.

You're looking at an old diff; I changed path.is_generic() to false in a later commit (the one we're nervous about ;).

I should probably remove the comment above this call too.

I have since dropped the commit that removed is_generic(), so this place now has resolved_path(..., path.is_assoc_ty()). I think resolved_path and related functions need to be cleaned up a lot in general, but I think let's do that in a different PR so this one doesn't keep growing ;)

Actually, it looks like I should be able to pretty easily clean up resolved_path such that the argument that path.is_generic() is used for is removed!

... or not; it looks like there's a single place where use_absolute is not false and not None :/

src/librustdoc/html/format.rs

src/librustdoc/clean/types.rs

src/librustdoc/html/render/print_item.rs

src/librustdoc/html/render/mod.rs

camelid · 2021-08-27T22:22:30Z

Got about halfway through and ran out of time - this is amazing work ❤️ not sure why it's marked as a draft, but feel free to ignore the comments until you're happy with it :)

Thank you! ❤️

It's mainly marked as draft because some of the changes need more investigation (e.g., #88379 (comment)).

camelid · 2021-08-27T22:25:21Z

Rebased to fix merge conflicts.

jyn514 · 2021-08-28T01:32:07Z

It's mainly marked as draft because some of the changes need more investigation (e.g., #88379 (comment)).

Ok. Can you split those into a separate PR so we can land the rest? There are a lot of great changes here :) but it's an awfully large diff and it makes it harder to review the more subtle changes.

camelid · 2021-08-28T01:34:24Z

It's mainly marked as draft because some of the changes need more investigation (e.g., #88379 (comment)).

Ok. Can you split those into a separate PR so we can land the rest? There are a lot of great changes here :) but it's an awfully large diff and it makes it harder to review the more subtle changes.

Sure! That's another reason why I marked it as draft, since I expected it would need to be split up.

camelid · 2021-08-28T23:48:33Z

@bors rollup=never

Since this may improve performance.

camelid · 2021-09-01T20:58:40Z

@jyn514 Ok, I think this should be ready for another review. There are a few unresolved threads:

I think the second and third ones should probably be left for future PRs. So, the first one is the main thread to be discussed. Or, if there's anything else you notice or have questions about, let me know!

I can also try to split up this PR further if it would make it easier to review.

src/librustdoc/clean/mod.rs

Instead, return `Type::Infer` since compilation should fail anyway. That's how rustdoc handles `hir::TyKind::Err`s, so this just extends that behavior to `ty::Err`s when analyzing associated types. For some reason, the error is printed twice with rustdoc (though only once with rustc). I'm not sure why that is, but it's better than panicking. This commit also makes rustdoc fail early in the non-projection, non-error case, instead of returning a `Res::Err` that would likely cause rustdoc to panic later on. This change is originally from rust-lang#88379.

bors · 2021-10-07T03:33:46Z

💥 Test timed out

rust-log-analyzer · 2021-10-07T03:45:38Z

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

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

camelid · 2021-10-07T04:11:01Z

I don't understand why this one builder (auto (dist-s390x-linux, ubuntu-latest-xl)) keeps timing out. This time it timed out after 4h24m, while even the macOS builder only took 2h47m. The build time for s390x-linux for the rollup PR that merged before my PR was tested was 2h 32m, so nowhere even close to the time out for my PR. The last part of the log is:

   Compiling tracing-tree v0.1.9
[RUSTC-TIMING] tracing_tree test:false 0.682
[RUSTC-TIMING] serde_derive test:false 5.011
[RUSTC-TIMING] serde_derive test:false 5.005
[RUSTC-TIMING] tracing_subscriber test:false 3.800
   Compiling rustdoc-json-types v0.1.0 (/checkout/src/rustdoc-json-types)
[RUSTC-TIMING] serde test:false 4.384
   Compiling tera v1.10.0
[RUSTC-TIMING] serde_json test:false 2.983
[RUSTC-TIMING] rustdoc_json_types test:false 2.985
   Compiling rustdoc v0.0.0 (/checkout/src/librustdoc)
[RUSTC-TIMING] tera test:false 24.695
Error: The operation was canceled.

So is it hanging while compiling rustdoc? Other PRs in the queue have merged, so it seems like it's specific to this PR. I just don't understand how my changes could cause only one of the builders to time out. Maybe monomorphization is taking longer since I removed a few Boxes? I still don't understand how it could be more than a negligible difference though.

EDIT: Based on the build log timestamps, it seems it is indeed timing out while compiling the rustdoc crate.

camelid · 2021-10-07T19:06:17Z

@hkratz suggests that this timeout may be due to #89609. Hopefully that'll be fixed soon, and this can get merged!

camelid · 2021-10-08T17:11:07Z

@bors retry

Stopgap CI fix is in a roll up at the front of the queue.

bors · 2021-10-09T00:08:30Z

⌛ Testing commit ebbcafb with merge 6eca024aa731feb4347f21453353a31e59ac09b9...

camelid · 2021-10-09T03:55:45Z

Looks like bors is likely to time out again, this time with i686-gnu. Although at least this time s390x-linux succeeded!

Since the last issue was LLVM related, I checked before/after for the update to the final LLVM 13.0 (the most recent change I saw to src/llvm-project), but CI actually got faster for i686-gnu after the update. It's still possible that's the cause, but it seems unlikely to me.

bors · 2021-10-09T04:23:30Z

💥 Test timed out

rust-log-analyzer · 2021-10-09T05:28:12Z

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

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

camelid · 2021-10-09T18:48:23Z

@ehuss suggested on Zulip that I retry this since it may be spurious.

@bors retry

bors · 2021-10-09T19:05:11Z

⌛ Testing commit ebbcafb with merge a8f2463...

bors · 2021-10-09T22:14:04Z

☀️ Test successful - checks-actions
Approved by: jyn514
Pushing a8f2463 to master...

rust-timer · 2021-10-09T23:46:54Z

Finished benchmarking commit (a8f2463): comparison url.

Summary: This change led to moderate relevant improvements 🎉 in compiler performance.

Moderate improvement in instruction counts (up to -0.8% on full builds of issue-46449)

If you disagree with this performance assessment, please file an issue in rust-lang/rustc-perf.

@rustbot label: -perf-regression

camelid added C-cleanup Category: PRs that clean code up or issues documenting cleanup. T-rustdoc Relevant to the rustdoc team, which will review and decide on the PR/issue. labels Aug 27, 2021

This comment has been minimized.

Sign in to view

jyn514 self-assigned this Aug 27, 2021

jyn514 reviewed Aug 27, 2021

View reviewed changes

camelid force-pushed the cleanup-clean branch from 9b36662 to 236a501 Compare August 27, 2021 22:25

camelid added the S-waiting-on-author Status: This is awaiting some action (such as code changes or more information) from the author. label Aug 27, 2021

camelid force-pushed the cleanup-clean branch from 236a501 to 973640c Compare August 27, 2021 23:34

This comment has been minimized.

Sign in to view

camelid force-pushed the cleanup-clean branch from 973640c to 1ae23d5 Compare August 27, 2021 23:41

camelid force-pushed the cleanup-clean branch 2 times, most recently from 96add5b to 6ecaa70 Compare August 28, 2021 02:14

camelid mentioned this pull request Aug 31, 2021

rustdoc: Box GenericArgs::Parenthesized.output #88522

Merged

This comment has been minimized.

Sign in to view

camelid force-pushed the cleanup-clean branch 3 times, most recently from c901187 to 1fca27b Compare September 1, 2021 18:44

camelid added S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. and removed S-waiting-on-author Status: This is awaiting some action (such as code changes or more information) from the author. labels Sep 1, 2021

camelid marked this pull request as ready for review September 1, 2021 20:59

camelid commented Sep 1, 2021

View reviewed changes

src/librustdoc/clean/mod.rs Outdated Show resolved Hide resolved

camelid mentioned this pull request Sep 1, 2021

rustdoc: Don't panic on ambiguous inherent associated types #88573

Merged

bors added S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. and removed S-waiting-on-bors Status: Waiting on bors to run and complete tests. Bors will change the label on completion. labels Oct 7, 2021

hkratz mentioned this pull request Oct 7, 2021

CI: dist-s390x-linux build went from 40min. to 160min with new LLVM pass manager #89609

Closed

bors added S-waiting-on-bors Status: Waiting on bors to run and complete tests. Bors will change the label on completion. and removed S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. labels Oct 8, 2021

camelid mentioned this pull request Oct 8, 2021

Default to disabling the new pass manager for the s390x arch targets. #89666

Merged

bors added S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. and removed S-waiting-on-bors Status: Waiting on bors to run and complete tests. Bors will change the label on completion. labels Oct 9, 2021

bors added S-waiting-on-bors Status: Waiting on bors to run and complete tests. Bors will change the label on completion. and removed S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. labels Oct 9, 2021

bors added the merged-by-bors This PR was explicitly merged by bors. label Oct 9, 2021

bors merged commit a8f2463 into rust-lang:master Oct 9, 2021

rustbot added this to the 1.57.0 milestone Oct 9, 2021

camelid deleted the cleanup-clean branch October 9, 2021 22:14

This was referenced Oct 9, 2021

Fix clippy lints in librustdoc #89430

Merged

rustdoc: Fix generics generation in search index #88268

Merged

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

rustdoc: Cleanup various `clean` types #88379

rustdoc: Cleanup various `clean` types #88379

camelid commented Aug 27, 2021 •

edited

Loading

This comment has been minimized.

jyn514 left a comment

jyn514 Aug 27, 2021

camelid Aug 27, 2021

camelid Sep 1, 2021 •

edited

Loading

jyn514 Aug 27, 2021

camelid Aug 27, 2021

camelid Aug 27, 2021

camelid Sep 1, 2021

camelid Sep 1, 2021

camelid Sep 1, 2021

camelid commented Aug 27, 2021

camelid commented Aug 27, 2021

This comment has been minimized.

jyn514 commented Aug 28, 2021

camelid commented Aug 28, 2021

camelid commented Aug 28, 2021

This comment has been minimized.

camelid commented Sep 1, 2021

bors commented Oct 7, 2021

rust-log-analyzer commented Oct 7, 2021

camelid commented Oct 7, 2021 •

edited

Loading

camelid commented Oct 7, 2021

camelid commented Oct 8, 2021

bors commented Oct 9, 2021

camelid commented Oct 9, 2021

bors commented Oct 9, 2021

rust-log-analyzer commented Oct 9, 2021

camelid commented Oct 9, 2021

bors commented Oct 9, 2021

bors commented Oct 9, 2021

rust-timer commented Oct 9, 2021

rustdoc: Cleanup various clean types #88379

rustdoc: Cleanup various clean types #88379

Conversation

camelid commented Aug 27, 2021 • edited Loading

This comment has been minimized.

jyn514 left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

camelid Sep 1, 2021 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

camelid commented Aug 27, 2021

camelid commented Aug 27, 2021

This comment has been minimized.

jyn514 commented Aug 28, 2021

camelid commented Aug 28, 2021

camelid commented Aug 28, 2021

This comment has been minimized.

camelid commented Sep 1, 2021

bors commented Oct 7, 2021

rust-log-analyzer commented Oct 7, 2021

camelid commented Oct 7, 2021 • edited Loading

camelid commented Oct 7, 2021

camelid commented Oct 8, 2021

bors commented Oct 9, 2021

camelid commented Oct 9, 2021

bors commented Oct 9, 2021

rust-log-analyzer commented Oct 9, 2021

camelid commented Oct 9, 2021

bors commented Oct 9, 2021

bors commented Oct 9, 2021

rust-timer commented Oct 9, 2021

rustdoc: Cleanup various `clean` types #88379

rustdoc: Cleanup various `clean` types #88379

camelid commented Aug 27, 2021 •

edited

Loading

camelid Sep 1, 2021 •

edited

Loading

camelid commented Oct 7, 2021 •

edited

Loading