sanity check field offsets in unsizeable structs

[lukas-code]

As promised in #112062 (comment), this PR extends the layout sanity checks to ensure that structs fields don't move around when unsizing and prevent issues like #112048 in the future. Like most other layout sanity checks, this only runs on compilers with debug assertions enabled.

Here is how it looks when it fails:

error: internal compiler error: compiler/rustc_ty_utils/src/layout.rs:533:21: unsizing GcNode<std::boxed::Box<i32>> changed field order!
                                Layout { size: Size(32 bytes), align: AbiAndPrefAlign { abi: Align(8 bytes), pref: Align(8 bytes) }, abi: Aggregate { sized: true }, fields: Arbitrary { offsets: [Size(0 bytes), Size(8 bytes), Size(24 bytes)], memory_index: [0, 1, 2] }, largest_niche: Some(Niche { offset: Size(24 bytes), value: Pointer(AddressSpace(0)), valid_range: 1..=18446744073709551615 }), variants: Single { index: 0 } }
                                Layout { size: Size(24 bytes), align: AbiAndPrefAlign { abi: Align(8 bytes), pref: Align(8 bytes) }, abi: Aggregate { sized: false }, fields: Arbitrary { offsets: [Size(16 bytes), Size(0 bytes), Size(24 bytes)], memory_index: [1, 0, 2] }, largest_niche: None, variants: Single { index: 0 } }

r? @the8472

[lukas-code]

@rustbot author

[lukas-code]

This will now pass always_sized = true to layout_of_struct_or_enum for enums and structs with no fields, but it shouldn't change behavior.

[lukas-code]

I didn't find a good solution for #113245 (comment). Instead I moved this check back to where the actual layout computation happens, so we can directly replace the tail type instead of messing with the generic parameters (substs) of the struct. This also means that replacing the tail will not change the other struct fields.

@rustbot ready

[lukas-code]

side note: The value passed for always_sized from rust-analyzer seems wrong, I think is_unsized should be is_sized instead.

rust/src/tools/rust-analyzer/crates/hir-ty/src/layout/adt.rs

Lines 102 to 107 in 131a036

	!matches!(def, AdtId::EnumId(..))
	&& variants
	.iter()
	.next()
	.and_then(|x| x.last().map(|x| x.is_unsized()))
	.unwrap_or(true),

[rustbot]

Some changes occurred to the core trait solver

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

[the8472]

@bors r+

[bors]

📌 Commit a28bb14982a9b469a998279ce0bc60a37c58d959 has been approved by the8472

It is now in the queue for this repository.

[the8472]

Hrmm, that test looks like it might be hitting the network and maybe the number of requests differ because it had to retry one, not sure.
Let's see if it's spurious.

@bors r+ rollup=iffy

[bors]

💡 This pull request was already approved, no need to approve it again.

• This pull request previously failed. You should add more commits to fix the bug, or use retry to trigger a build again.
• There's another pull request that is currently being tested, blocking this pull request: Remove some unnecessary(?) normalization #113348

[bors]

📌 Commit a28bb14982a9b469a998279ce0bc60a37c58d959 has been approved by the8472

It is now in the queue for this repository.

[bors]

⌛ Testing commit a28bb14982a9b469a998279ce0bc60a37c58d959 with merge e3f0b86d0879138b8f4e8b755205c729f45870fc...

[bors]

💔 Test failed - checks-actions

[lukas-code]

rebased onto #113377

[the8472]

@bors retry

[lukas-code]

I don't see this PR in the queue.

I think you need to r+ again, because I pushed new commits.

[the8472]

@bors r+

[bors]

📌 Commit 7aa5f39 has been approved by the8472

It is now in the queue for this repository.

[bors]

⌛ Testing commit 7aa5f39 with merge cb80ff1...

[bors]

☀️ Test successful - checks-actions
Approved by: the8472
Pushing cb80ff1 to master...

[rust-timer]

Finished benchmarking commit (cb80ff1): comparison URL.

Overall result: ❌✅ regressions and 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.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	1.6%	[1.6%, 1.6%]	1
Improvements ✅ (primary)	-0.6%	[-0.6%, -0.6%]	1
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-0.6%	[-0.6%, -0.6%]	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)	2.2%	[2.2%, 2.2%]	1
Regressions ❌ (secondary)	2.7%	[2.5%, 2.9%]	2
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	2.2%	[2.2%, 2.2%]	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)	3.1%	[2.4%, 4.2%]	7
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	3.1%	[2.4%, 4.2%]	7

Binary size

This benchmark run did not return any relevant results for this metric.

Bootstrap: 657.311s -> 655.85s (-0.22%)