no unnormalized types for implied bounds outside borrowck

[lcnr]

fixes #100910 - introduced in #100676 - by only considering normalized types for wf.

r? types

[lcnr]

@bors rollup=never

[spastorino]

@bors r+

[bors]

📌 Commit ef7bda3 has been approved by spastorino

It is now in the queue for this repository.

[jackh726]

Why does this not break the tests that rely on the implied bounds of unnormalized types?

It kind of concerns me that we don't have test coverage here. Either this line is important for functionality, and there should be some test that breaks. Or it doesn't do anything and it never needs to be added back in. It's not just a bug fix.

[jackh726]

I'm not going to r- this, but I think we need to get a regression test here that fails with this PR. And also be on the lookout for but reports that might be caused by this.

[lcnr]

Why does this not break the tests that rely on the implied bounds of unnormalized types?

implied bounds are currently computed in 2 different ways depending on whether we're in borrowck or wfcheck. Borrowck was changed to consider unnormalized types (and using functions correctly adds unnormalized wf obligations), but wfcheck didn't use unnormalized types before #100676 and that PR didn't add any tests for that as I somehow missed that while implementing it.

When adding the unnormalized types back to the implied bounds outside of borrowck, we definitely need some tests showing that this is the case and that we correctly check the unnormalized type for wf

[bors]

⌛ Testing commit ef7bda3 with merge 3b3f3b7...

[bors]

☀️ Test successful - checks-actions
Approved by: spastorino
Pushing 3b3f3b7 to master...

[rust-timer]

Finished benchmarking commit (3b3f3b7): comparison URL.

Overall result: ✅ improvements - no action needed

@rustbot label: -perf-regression

Instruction count

This is a highly reliable metric that was used to determine the overall result at the top of this comment.

	mean1	range	count2
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-1.2%	[-1.2%, -1.2%]	2
All ❌✅ (primary)	-	-	0

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.

	mean1	range	count2
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-4.3%	[-4.3%, -4.3%]	1
Improvements ✅ (secondary)	-2.5%	[-3.1%, -1.4%]	3
All ❌✅ (primary)	-4.3%	[-4.3%, -4.3%]	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.

	mean1	range	count2
Regressions ❌ (primary)	2.2%	[2.2%, 2.2%]	1
Regressions ❌ (secondary)	3.6%	[3.6%, 3.6%]	1
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	2.2%	[2.2%, 2.2%]	1

Footnotes

1. the arithmetic mean of the percent change ↩ ↩² ↩³

2. number of relevant changes ↩ ↩² ↩³