Do not unify dereferences of shared borrows in GVN

[RalfJung]

Repost of #132461, the last commit applies my suggestions.

Fixes #130853

[rustbot]

r? @Nadrieril

rustbot has assigned @Nadrieril.
They will have a look at your PR within the next two weeks and either review your PR or reassign to another reviewer.

Use r? to explicitly pick a reviewer

[rustbot]

Some changes occurred to MIR optimizations

cc @rust-lang/wg-mir-opt

Some changes occurred in coverage tests.

cc @Zalathar

[RalfJung]

@bors try
@rust-timer queue

[bors]

⌛ Trying commit 906f66f with merge 16c1019...

[compiler-errors]

r=me after try build is done :D

[bors]

☀️ Try build successful - checks-actions
Build commit: 16c1019 (16c101943b84f24b28a1a713b06e961002c2440b)

[rust-timer]

Finished benchmarking commit (16c1019): 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)	0.3%	[0.2%, 0.7%]	8
Regressions ❌ (secondary)	0.2%	[0.1%, 0.4%]	13
Improvements ✅ (primary)	-1.9%	[-6.7%, -0.3%]	8
Improvements ✅ (secondary)	-0.2%	[-0.3%, -0.1%]	5
All ❌✅ (primary)	-0.8%	[-6.7%, 0.7%]	16

Max RSS (memory usage)

Results (primary -1.4%, secondary 3.4%)

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)	2.8%	[1.2%, 4.1%]	3
Regressions ❌ (secondary)	3.4%	[3.4%, 3.4%]	1
Improvements ✅ (primary)	-5.6%	[-6.6%, -4.1%]	3
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-1.4%	[-6.6%, 4.1%]	6

Cycles

Results (primary -3.4%)

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)	1.5%	[1.5%, 1.5%]	1
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-5.9%	[-6.5%, -5.2%]	2
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-3.4%	[-6.5%, 1.5%]	3

Binary size

Results (primary 0.0%, secondary 0.0%)

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)	0.3%	[0.0%, 2.6%]	32
Regressions ❌ (secondary)	0.5%	[0.1%, 1.1%]	7
Improvements ✅ (primary)	-0.3%	[-2.0%, -0.0%]	28
Improvements ✅ (secondary)	-0.1%	[-0.3%, -0.0%]	41
All ❌✅ (primary)	0.0%	[-2.0%, 2.6%]	60

Bootstrap: 796.592s -> 797.482s (0.11%)
Artifact size: 336.27 MiB -> 336.00 MiB (-0.08%)

[RalfJung]

@bors r=compiler-errors

[bors]

📌 Commit 906f66f has been approved by compiler-errors

It is now in the queue for this repository.

[bors]

⌛ Testing commit 906f66f with merge 6b6a867...

[bors]

☀️ Test successful - checks-actions
Approved by: compiler-errors
Pushing 6b6a867 to master...

[rust-timer]

Finished benchmarking commit (6b6a867): comparison URL.

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

Our benchmarks found a performance regression caused by this PR.
This might be an actual regression, but it can also be just noise.

Next Steps:

• If the regression was expected or you think it can be justified,
  please write a comment with sufficient written justification, and add
  @rustbot label: +perf-regression-triaged to it, to mark the regression as triaged.
• If you think that you know of a way to resolve the regression, try to create
  a new PR with a fix for the regression.
• If you do not understand the regression or you think that it is just noise,
  you can ask the @rust-lang/wg-compiler-performance working group for help (members of this group
  were already notified of this PR).

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

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)	0.2%	[0.1%, 0.6%]	14
Regressions ❌ (secondary)	0.2%	[0.1%, 0.4%]	10
Improvements ✅ (primary)	-1.7%	[-6.7%, -0.2%]	9
Improvements ✅ (secondary)	-0.2%	[-0.3%, -0.1%]	6
All ❌✅ (primary)	-0.5%	[-6.7%, 0.6%]	23

Max RSS (memory usage)

Results (primary 2.0%)

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)	5.3%	[3.1%, 7.3%]	4
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-4.4%	[-4.6%, -4.2%]	2
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	2.0%	[-4.6%, 7.3%]	6

Cycles

Results (primary -3.6%, secondary 2.5%)

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)	1.7%	[1.7%, 1.7%]	1
Regressions ❌ (secondary)	2.5%	[2.5%, 2.5%]	1
Improvements ✅ (primary)	-6.2%	[-6.7%, -5.7%]	2
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-3.6%	[-6.7%, 1.7%]	3

Binary size

Results (primary -0.0%, secondary 0.0%)

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)	0.3%	[0.0%, 2.6%]	28
Regressions ❌ (secondary)	0.5%	[0.1%, 1.1%]	7
Improvements ✅ (primary)	-0.3%	[-2.1%, -0.0%]	28
Improvements ✅ (secondary)	-0.1%	[-0.3%, -0.0%]	41
All ❌✅ (primary)	-0.0%	[-2.1%, 2.6%]	56

Bootstrap: 795.759s -> 793.974s (-0.22%)
Artifact size: 336.16 MiB -> 335.91 MiB (-0.07%)

[rylev]

@RalfJung I'm doing perf triage and noticed you ran perf, had regressions, but didn't come on why this regressions are acceptable in this case. Could you indicate why the regression here is acceptable?

[compiler-errors]

@rylev: this fixes a miscompilation due to an unsound MIR optimization which is disabled in this PR.

[rylev]

Figured that the answer was "correctness over performance", but I just wanted to double check 😄

@rustbot label: +perf-regression-triaged

[RalfJung]

Well also overall this was an improvement, no?

[Kobzol]

The regex win seems genuine, the html5ever was perhaps only codegen noise, as it bumped back soon after (in #123244, which had a large perf. effect though, so it could have been truly caused by that PR).

Numerically it was a win, yeah :)