Prototype: Add unstable `-Z reference-niches` option #113166

moulins · 2023-06-29T17:46:42Z

MCP: rust-lang/compiler-team#641
Relevant RFC: rust-lang/rfcs#3204

This prototype adds a new -Z reference-niches option, controlling the range of valid bit-patterns for reference types (&T and &mut T), thereby enabling new enum niching opportunities. Like -Z randomize-layout, this setting is crate-local; as such, references to built-in types (primitives, tuples, ...) are not affected.

The possible settings are (here, MAX denotes the all-1 bit-pattern):

`-Z reference-niches=`	Valid range
`null` (the default)	`1..=MAX`
`size`	`1..=(MAX- size)`
`align`	`align..=MAX.align_down_to(align)`
`size,align`	`align..=(MAX-size).align_down_to(align)`

This is very WIP, and I'm not sure the approach I've taken here is the best one, but stage 1 tests pass locally; I believe this is in a good enough state to unleash this upon unsuspecting 3rd-party code, and see what breaks.

rustbot · 2023-06-29T17:47:58Z

Thanks for the pull request, and welcome! The Rust team is excited to review your changes, and you should hear from @jackh726 (or someone else) soon.

Please see the contribution instructions for more information. Namely, in order to ensure the minimum review times lag, PR authors and assigned reviewers should ensure that the review label (S-waiting-on-review and S-waiting-on-author) stays updated, invoking these commands when appropriate:

@rustbot author: the review is finished, PR author should check the comments and take action accordingly
@rustbot review: the author is ready for a review, this PR will be queued again in the reviewer's queue

rustbot · 2023-06-29T17:50:37Z

Some changes occurred to the CTFE / Miri engine

cc @rust-lang/miri

Some changes occurred to the CTFE / Miri engine

cc @rust-lang/miri

Some changes occurred to the core trait solver

cc @rust-lang/initiative-trait-system-refactor

oli-obk · 2023-06-29T18:38:48Z

r? @oli-obk

moulins · 2023-06-29T23:51:16Z

This should fix the stage 2 rustc build.

compiler/rustc_abi/src/lib.rs

RalfJung · 2023-06-30T10:14:46Z

This prototype adds a new -Z reference-niches option, controlling the range of valid bit-patterns for reference types (&T and &mut T), thereby enabling new enum niching opportunities. Like -Z randomize-layout, this setting is crate-local; as such, references to built-in types (primitives, tuples, ...) are not affected.

I am confused by this comment. Does this mean that it is the crate where T was defined that decides whether &T gets the layout optimization? That's not how the docs sound like.

Also does this mean &i32 never gets the new niches?

compiler/rustc_const_eval/src/interpret/discriminant.rs

compiler/rustc_const_eval/src/interpret/memory.rs

compiler/rustc_abi/src/lib.rs

tests/ui/lint/invalid_value.stderr

compiler/rustc_middle/src/query/mod.rs

oli-obk

Please rewrite your git history so that every commit that affects any tests, has its tests blessed, so we can track test changes as the commits progress.

compiler/rustc_middle/src/ty/layout.rs

compiler/rustc_ty_utils/src/layout.rs

compiler/rustc_middle/src/ty/layout.rs

oli-obk · 2023-06-30T13:09:05Z

I had another comment that github ate and then I forgot it. Either I'll remember in a second round of reviewing or it's not that important.

This is really cool!

@bors try @rust-timer queue (let's see that none of the changes affect perf even with the feature gate disabled)

bors · 2023-06-30T13:09:14Z

⌛ Trying commit bffc832610817ff707c591994d02864d3225263c with merge 43dda3cbf8a740d44e87bca311d4827e17b2ab4e...

bors · 2023-06-30T14:45:50Z

☀️ Try build successful - checks-actions
Build commit: 43dda3cbf8a740d44e87bca311d4827e17b2ab4e (43dda3cbf8a740d44e87bca311d4827e17b2ab4e)

rust-timer · 2023-06-30T17:05:12Z

Finished benchmarking commit (43dda3cbf8a740d44e87bca311d4827e17b2ab4e): comparison URL.

Overall result: ❌ regressions - 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)	0.9%	[0.7%, 0.9%]	5
Regressions ❌ (secondary)	1.2%	[1.1%, 1.3%]	3
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	0.9%	[0.7%, 0.9%]	5

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)	1.8%	[0.4%, 4.4%]	23
Regressions ❌ (secondary)	2.4%	[2.0%, 3.4%]	9
Improvements ✅ (primary)	-2.9%	[-2.9%, -2.9%]	1
Improvements ✅ (secondary)	-2.2%	[-4.6%, -1.0%]	3
All ❌✅ (primary)	1.6%	[-2.9%, 4.4%]	24

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)	2.1%	[2.1%, 2.1%]	2
Regressions ❌ (secondary)	1.8%	[1.8%, 1.8%]	1
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	2.1%	[2.1%, 2.1%]	2

Binary size

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)	0.0%	[0.0%, 0.0%]	1
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-	-	0

Bootstrap: 659.918s -> 667.792s (1.19%)

Still more complexity, but this allows computing exact `NaiveLayout`s for null-optimized enums, and thus allows calls like `transmute::<Option<&T>, &U>()` to work in generic contexts.

oli-obk · 2023-07-21T13:17:11Z

@bors r+

bors · 2023-07-21T13:17:13Z

📌 Commit 7f10908 has been approved by oli-obk

It is now in the queue for this repository.

bors · 2023-07-21T15:00:40Z

⌛ Testing commit 7f10908 with merge 557359f...

bors · 2023-07-21T16:52:14Z

☀️ Test successful - checks-actions
Approved by: oli-obk
Pushing 557359f to master...

rust-timer · 2023-07-21T18:56:36Z

Finished benchmarking commit (557359f): comparison URL.

Overall result: ❌ regressions - ACTION NEEDED

Next Steps: If you can justify the regressions found in this perf run, please indicate this with @rustbot label: +perf-regression-triaged along with sufficient written justification. If you cannot justify the regressions please open an issue or create a new PR that fixes the regressions, add a comment linking to the newly created issue or PR, and then add the perf-regression-triaged label to this PR.

@rustbot label: +perf-regression
cc @rust-lang/wg-compiler-performance

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)	0.7%	[0.3%, 1.1%]	19
Regressions ❌ (secondary)	1.0%	[0.3%, 1.2%]	4
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	0.7%	[0.3%, 1.1%]	19

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)	2.3%	[0.9%, 4.3%]	13
Regressions ❌ (secondary)	2.2%	[1.8%, 2.9%]	7
Improvements ✅ (primary)	-2.2%	[-2.2%, -2.2%]	1
Improvements ✅ (secondary)	-3.4%	[-5.9%, -2.3%]	4
All ❌✅ (primary)	2.0%	[-2.2%, 4.3%]	14

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)	3.0%	[1.0%, 4.2%]	10
Regressions ❌ (secondary)	2.1%	[2.1%, 2.1%]	1
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	3.0%	[1.0%, 4.2%]	10

Binary size

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)	-0.2%	[-0.2%, -0.2%]	1
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-0.2%	[-0.2%, -0.2%]	1

Bootstrap: 652.363s -> 658.335s (0.92%)

…r=oli-obk" This reverts commit 557359f, reversing changes made to 1e6c09a.

Revert "Prototype: Add unstable `-Z reference-niches` option" Clean revert of rust-lang#113166. I confirmed this fixes rust-lang#113941.

pnkfelix · 2023-07-26T16:21:23Z

reverted by PR #113946, which restored the performance, at least for the truly significant regressive cases.

@rustbot label: +perf-regression-triaged

Prototype: Add unstable `-Z reference-niches` option MCP: rust-lang/compiler-team#641 Relevant RFC: rust-lang/rfcs#3204 This prototype adds a new `-Z reference-niches` option, controlling the range of valid bit-patterns for reference types (`&T` and `&mut T`), thereby enabling new enum niching opportunities. Like `-Z randomize-layout`, this setting is crate-local; as such, references to built-in types (primitives, tuples, ...) are not affected. The possible settings are (here, `MAX` denotes the all-1 bit-pattern): | `-Z reference-niches=` | Valid range | |:---:|:---:| | `null` (the default) | `1..=MAX` | | `size` | `1..=(MAX- size)` | | `align` | `align..=MAX.align_down_to(align)` | | `size,align` | `align..=(MAX-size).align_down_to(align)` | ------ This is very WIP, and I'm not sure the approach I've taken here is the best one, but stage 1 tests pass locally; I believe this is in a good enough state to unleash this upon unsuspecting 3rd-party code, and see what breaks.

…r=oli-obk" This reverts commit 557359f, reversing changes made to 1e6c09a.

Add range metadata to slice lengths This adds range information to the slice-len in fat pointers if we can conservatively determine that the pointee is not a ZST without having to normalize the pointee type. I only intended to pass the `!range` to llvm but apparently this also lets the length in fat pointers be used for its niches 😅. Ideally this would use the naive-layout computation from rust-lang#113166 to calculate a better approximation of the pointee size, but that PR got reverted.

moulins marked this pull request as ready for review June 29, 2023 17:50

This comment has been minimized.

Sign in to view

rustbot assigned oli-obk and unassigned jackh726 Jun 29, 2023

moulins force-pushed the ref-niches-initial branch from 716cd2c to bffc832 Compare June 29, 2023 23:50

This comment has been minimized.

Sign in to view

RalfJung reviewed Jun 30, 2023

View reviewed changes

compiler/rustc_abi/src/lib.rs Show resolved Hide resolved

RalfJung reviewed Jun 30, 2023

View reviewed changes

compiler/rustc_abi/src/lib.rs Outdated Show resolved Hide resolved

RalfJung reviewed Jun 30, 2023

View reviewed changes

oli-obk reviewed Jun 30, 2023

View reviewed changes

compiler/rustc_middle/src/query/mod.rs Outdated Show resolved Hide resolved

oli-obk requested changes Jun 30, 2023

View reviewed changes

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 Jun 30, 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 Jun 30, 2023

Track (partial) niche information in NaiveLayout

7f10908

Still more complexity, but this allows computing exact `NaiveLayout`s for null-optimized enums, and thus allows calls like `transmute::<Option<&T>, &U>()` to work in generic contexts.

moulins force-pushed the ref-niches-initial branch from 3f0cf51 to 7f10908 Compare July 21, 2023 12:26

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-author Status: This is awaiting some action (such as code changes or more information) from the author. labels Jul 21, 2023

bors added the merged-by-bors This PR was explicitly merged by bors. label Jul 21, 2023

bors merged commit 557359f into rust-lang:master Jul 21, 2023

rustbot added this to the 1.73.0 milestone Jul 21, 2023

This was referenced Jul 21, 2023

Default relax_elf_relocations to true #106511

Merged

Compute NLL loan scopes using the polonius model #113218

Merged

Refactor const_eval diagnostics #113297

Closed

Don't attempt to compute layout of type referencing error #113773

Merged

This was referenced Jul 22, 2023

the naive layout isn't refined by the actual layout #113941

Closed

Revert "Prototype: Add unstable -Z reference-niches option" #113946

Merged

dtolnay added a commit to dtolnay/rust that referenced this pull request Jul 22, 2023

Revert "Auto merge of rust-lang#113166 - moulins:ref-niches-initial, …

5bbf0a8

…r=oli-obk" This reverts commit 557359f, reversing changes made to 1e6c09a.

matthiaskrgr mentioned this pull request Jul 22, 2023

ICE: the naive layout isn't refined by the actual layout #113952

Closed

rustbot added the perf-regression-triaged The performance regression has been triaged. label Jul 26, 2023

RalfJung mentioned this pull request Aug 1, 2023

It should be possible to get a mutable reference to a ZST literal #103821

Open

the8472 mentioned this pull request Oct 8, 2023

Add range metadata to slice lengths #116542

Draft

antoyo pushed a commit to antoyo/rust that referenced this pull request Oct 9, 2023

Revert "Auto merge of rust-lang#113166 - moulins:ref-niches-initial, …

e2a8474

…r=oli-obk" This reverts commit 557359f, reversing changes made to 1e6c09a.

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Prototype: Add unstable `-Z reference-niches` option #113166

Prototype: Add unstable `-Z reference-niches` option #113166

moulins commented Jun 29, 2023

rustbot commented Jun 29, 2023

rustbot commented Jun 29, 2023

This comment has been minimized.

oli-obk commented Jun 29, 2023

moulins commented Jun 29, 2023

This comment has been minimized.

RalfJung commented Jun 30, 2023 •

edited

Loading

oli-obk left a comment

oli-obk commented Jun 30, 2023

This comment has been minimized.

bors commented Jun 30, 2023

bors commented Jun 30, 2023

This comment has been minimized.

rust-timer commented Jun 30, 2023

oli-obk commented Jul 21, 2023

bors commented Jul 21, 2023

bors commented Jul 21, 2023

bors commented Jul 21, 2023

rust-timer commented Jul 21, 2023

pnkfelix commented Jul 26, 2023

Prototype: Add unstable -Z reference-niches option #113166

Prototype: Add unstable -Z reference-niches option #113166

Conversation

moulins commented Jun 29, 2023

rustbot commented Jun 29, 2023

rustbot commented Jun 29, 2023

This comment has been minimized.

oli-obk commented Jun 29, 2023

moulins commented Jun 29, 2023

This comment has been minimized.

RalfJung commented Jun 30, 2023 • edited Loading

oli-obk left a comment

Choose a reason for hiding this comment

oli-obk commented Jun 30, 2023

This comment has been minimized.

bors commented Jun 30, 2023

bors commented Jun 30, 2023

This comment has been minimized.

rust-timer commented Jun 30, 2023

Overall result: ❌ regressions - ACTION NEEDED

oli-obk commented Jul 21, 2023

bors commented Jul 21, 2023

bors commented Jul 21, 2023

bors commented Jul 21, 2023

rust-timer commented Jul 21, 2023

Overall result: ❌ regressions - ACTION NEEDED

pnkfelix commented Jul 26, 2023

Prototype: Add unstable `-Z reference-niches` option #113166

Prototype: Add unstable `-Z reference-niches` option #113166

RalfJung commented Jun 30, 2023 •

edited

Loading