[experimental, do not merge!] a faster implementation of Polonius and a more compact DenseBitSet implementation

[tage64]

This is the union of #141326 and #141325, a Polonius experiment combined with a more compact version of DenseBitSet. I would like to get a perf-run of this @lqd.

r? lqd

[rustbot]

This PR changes a file inside tests/crashes. If a crash was fixed, please move into the corresponding ui subdir and add 'Fixes #' to the PR description to autoclose the issue upon merge.

Some changes occurred in src/tools/compiletest

cc @jieyouxu

Some changes occurred to MIR optimizations

cc @rust-lang/wg-mir-opt

These commits modify the Cargo.lock file. Unintentional changes to Cargo.lock can be introduced when switching branches and rebasing PRs.

If this was unintentional then you should revert the changes before this PR is merged.
Otherwise, you can ignore this comment.

Some changes occurred in coverage instrumentation.

cc @Zalathar

[lqd]

For perf tests, your PRs should be drafts so they’re not merged and also they won’t ping a lot of people until you want them to do so for review.

@bors try @rust-timer queue

[bors]

⌛ Trying commit f2585eb with merge 16435a6...

[bors]

☀️ Try build successful - checks-actions
Build commit: 16435a6 (16435a66301a7d82ae6ff183f6d9a7005cb3f471)

[Kobzol]

For perf tests, your PRs should be drafts so they’re not merged and also they won’t ping a lot of people until you want them to do so for review.

I think that drafts still ping ppl right now, although they shouldn't request a reviewer.

[lqd]

Sometimes we don't want to ping people, so I hope that's incorrect and they still don't ping.

[rust-timer]

Finished benchmarking commit (16435a6): comparison URL.

Overall result: ❌✅ regressions and improvements - please read the text below

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 the most reliable metric that we have; it was used to determine the overall result at the top of this comment. However, even this metric can sometimes exhibit noise.

	mean	range	count
Regressions ❌ (primary)	2.1%	[0.1%, 5.1%]	7
Regressions ❌ (secondary)	1.4%	[0.3%, 3.6%]	12
Improvements ✅ (primary)	-0.5%	[-3.0%, -0.1%]	69
Improvements ✅ (secondary)	-1.0%	[-2.9%, -0.1%]	47
All ❌✅ (primary)	-0.3%	[-3.0%, 5.1%]	76

Max RSS (memory usage)

Results (primary 70.8%, secondary 25.3%)

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)	76.9%	[0.9%, 191.0%]	12
Regressions ❌ (secondary)	25.3%	[2.3%, 44.2%]	8
Improvements ✅ (primary)	-2.2%	[-2.2%, -2.2%]	1
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	70.8%	[-2.2%, 191.0%]	13

Cycles

Results (primary 1.4%, secondary 1.0%)

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)	2.4%	[1.0%, 3.6%]	16
Regressions ❌ (secondary)	2.5%	[1.9%, 3.3%]	5
Improvements ✅ (primary)	-1.6%	[-2.8%, -0.9%]	5
Improvements ✅ (secondary)	-1.6%	[-2.0%, -1.3%]	3
All ❌✅ (primary)	1.4%	[-2.8%, 3.6%]	21

Binary size

Results (primary 0.0%, secondary 0.1%)

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)	0.0%	[0.0%, 0.1%]	46
Regressions ❌ (secondary)	0.1%	[0.0%, 0.1%]	29
Improvements ✅ (primary)	-1.1%	[-1.1%, -1.1%]	1
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	0.0%	[-1.1%, 0.1%]	47

Bootstrap: 776.359s -> 774.569s (-0.23%)
Artifact size: 366.28 MiB -> 366.23 MiB (-0.02%)

[lqd]

cranelift-codegen is still seeing a big +5% hit but funnily enough I likely have a -10% improvement for it.

I haven't looked into the code yet apart from our discussions and walkthrough, are the max-rss increases expected?

[tage64]

cranelift-codegen is still seeing a big +5% hit but funnily enough I likely have a -10% improvement for it.

Great! How do you get that improvement? In the borrow checker, or somewhere else?

I haven't looked into the code yet apart from our discussions and walkthrough, are the max-rss increases expected?

No The max-rss increase in a few crates is not expected. I don't know what it comes from. It is at least not the DenseBitSet implementation because we see the same regression in #141326 which is same as this pr minus the DenseBitSet-changes. It has to be investigated further.

[lqd]

In move/init for liveness in the borrowck, #141667