[PERF] allow region unification to change region universes by aliemjay · Pull Request #108867 · rust-lang/rust

aliemjay · 2023-03-07T16:03:26Z

forked from #108121.

compiler-errors · 2023-03-08T03:35:53Z

bors · 2023-03-08T03:36:06Z

⌛ Trying commit 4cb44a7 with merge 1bced552f1758b8f979780b55c326f0a39c51181...

bors · 2023-03-08T05:47:12Z

☀️ Try build successful - checks-actions
Build commit: 1bced552f1758b8f979780b55c326f0a39c51181 (1bced552f1758b8f979780b55c326f0a39c51181)

rust-timer · 2023-03-08T07:03:56Z

Finished benchmarking commit (1bced552f1758b8f979780b55c326f0a39c51181): comparison URL.

Overall result: ❌✅ regressions and improvements - ACTION NEEDED

Benchmarking this pull request likely means that it is perf-sensitive, so we're automatically marking it as not fit for rolling up. While you can manually mark this PR as fit for rollup, we strongly recommend not doing so since this PR may lead to changes in compiler perf.

Next Steps: If you can justify the regressions found in this try perf run, please indicate this with @rustbot label: +perf-regression-triaged along with sufficient written justification. If you cannot justify the regressions please fix the regressions and do another perf run. If the next run shows neutral or positive results, the label will be automatically removed.

@bors rollup=never
@rustbot label: -S-waiting-on-perf +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)	0.7%	[0.7%, 0.8%]	2
Improvements ✅ (primary)	-0.2%	[-0.2%, -0.2%]	1
Improvements ✅ (secondary)	-3.1%	[-4.2%, -0.6%]	9
All ❌✅ (primary)	-0.2%	[-0.2%, -0.2%]	1

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)	-4.1%	[-4.1%, -4.1%]	1
Improvements ✅ (secondary)	-6.4%	[-8.4%, -5.3%]	6
All ❌✅ (primary)	-4.1%	[-4.1%, -4.1%]	1

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)	-6.2%	[-6.5%, -5.9%]	6
All ❌✅ (primary)	-	-	0

lcnr

unhappy to have the universe both in the var_infos and in the inference data. I don't think we should error eagerly for universe errors when equating universes because the new solver erases the universe of regions in the argument. THis makes adding universe to the UnifiedRegion really cumbersome I think. It's different for https://doc.rust-lang.org/nightly/nightly-rustc/rustc_infer/infer/type_variable/enum.TypeVariableValue.html and consts.

Unless you now prefer this approach, I would go ahead with #108121. Thank you for actually looking into my suggestion even though you disagreed ❤️

lcnr · 2023-03-08T07:26:20Z

compiler/rustc_middle/src/infer/unify_key.rs

+pub struct UnifiedRegion<'tcx> {
+    pub universe: ty::UniverseIndex,
+    pub value: Option<ty::Region<'tcx>>,
+}


similar to TypeVariableValue and ConstVariableValue change this to an enum and only have a universe in the Unknown case? 🤔

although it seems redundant, but I thought it might be a good idea to store the universe of universals and existentials in the same place.

if region unification is infallible i think we even have to keep the universe even if value is Some, don't we? 🤔

I don't think so. universe field here is redundant if value is Some.

lcnr · 2023-03-08T07:37:51Z

compiler/rustc_middle/src/infer/unify_key.rs

-            (None, None) => *value1,
-        })
+    fn unify_values(value1: &Self, value2: &Self) -> Result<Self, Self::Error> {
+        Ok(cmp::min_by_key(*value1, *value2, |val| (val.universe, val.value.is_none())))


this can pull down the universe of known regions which seems wrong 🤔 I would expect us to have to eagerly error here if we relate with a known region

The idea is that unification should always be infallible, the constraint graph will eventually emit a universe error in this case.

The only invariant we should maintain in the unification table is that an existential cannot resolve to a region in a higher universe (plus one nice thing to have is to prefer universals over existentials within the same universe, but that's not necessary for soundness).

aliemjay · 2023-03-08T09:14:56Z

Unless you now prefer this approach, I would go ahead with #108121. Thank you for actually looking into my suggestion even though you disagreed heart

Indeed I prefer the approach :) but I'll go ahead and land #108121 first to unblock my other PRs and we'll get back to this later.

jyn514 · 2023-03-08T15:35:14Z

@craterbot run mode=check-only

craterbot · 2023-03-08T15:35:43Z

👌 Experiment pr-108867 created and queued.
🤖 Automatically detected try build 1bced552f1758b8f979780b55c326f0a39c51181
🔍 You can check out the queue and this experiment's details.

ℹ️ Crater is a tool to run experiments across parts of the Rust ecosystem. Learn more

craterbot · 2023-03-08T15:36:14Z

🚧 Experiment pr-108867 is now running

ℹ️ Crater is a tool to run experiments across parts of the Rust ecosystem. Learn more

craterbot · 2023-03-09T13:22:42Z

🎉 Experiment pr-108867 is completed!
📊 15 regressed and 1 fixed (257954 total)
📰 Open the full report.

⚠️ If you notice any spurious failure please add them to the blacklist!
ℹ️ Crater is a tool to run experiments across parts of the Rust ecosystem. Learn more

Dylan-DPC · 2023-08-07T13:57:53Z

closing this as it was a perf experiment

aliemjay added 3 commits March 6, 2023 19:24

account for universes in region unification

7e9b2ab

make unification infallible again

dc57c8b

bless rustdoc tests

a94140f

rustbot added S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. T-compiler Relevant to the compiler 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. labels Mar 7, 2023

aliemjay mentioned this pull request Mar 7, 2023

always resolve to universal regions if possible #108121

Merged

pacify tidy

4cb44a7

This comment has been minimized.

Sign in to view

rustbot added the S-waiting-on-perf Status: Waiting on a perf run to be completed. label Mar 8, 2023

This comment has been minimized.

Sign in to view

rustbot added perf-regression Performance regression. and removed S-waiting-on-perf Status: Waiting on a perf run to be completed. labels Mar 8, 2023

lcnr reviewed Mar 8, 2023

View reviewed changes

craterbot added S-waiting-on-crater Status: Waiting on a crater run to be completed. and removed S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. labels Mar 8, 2023

craterbot added S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. and removed S-waiting-on-crater Status: Waiting on a crater run to be completed. labels Mar 9, 2023

lcnr added S-waiting-on-author Status: This is awaiting some action (such as code changes or more information) from the author. and removed S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. labels Mar 10, 2023

Dylan-DPC closed this Aug 7, 2023

Dylan-DPC added S-experimental Status: Ongoing experiment that does not require reviewing and won't be merged in its current state. and removed S-waiting-on-author Status: This is awaiting some action (such as code changes or more information) from the author. labels Aug 7, 2023

Uh oh!

Conversation

aliemjay commented Mar 7, 2023

Uh oh!

compiler-errors commented Mar 8, 2023

Uh oh!

This comment has been minimized.

bors commented Mar 8, 2023

Uh oh!

bors commented Mar 8, 2023

Uh oh!

This comment has been minimized.

rust-timer commented Mar 8, 2023

Overall result: ❌✅ regressions and improvements - ACTION NEEDED

Instruction count

Max RSS (memory usage)

Cycles

Uh oh!

lcnr left a comment • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

lcnr Mar 8, 2023

Choose a reason for hiding this comment

Uh oh!

aliemjay Mar 8, 2023 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

lcnr Mar 8, 2023

Choose a reason for hiding this comment

Uh oh!

aliemjay Mar 8, 2023

Choose a reason for hiding this comment

Uh oh!

lcnr Mar 8, 2023

Choose a reason for hiding this comment

Uh oh!

aliemjay Mar 8, 2023 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

aliemjay commented Mar 8, 2023

Uh oh!

jyn514 commented Mar 8, 2023

Uh oh!

craterbot commented Mar 8, 2023

Uh oh!

craterbot commented Mar 8, 2023

Uh oh!

craterbot commented Mar 9, 2023

Uh oh!

Dylan-DPC commented Aug 7, 2023

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

9 participants

lcnr left a comment •

edited

Loading

aliemjay Mar 8, 2023 •

edited

Loading

aliemjay Mar 8, 2023 •

edited

Loading