Make the randomize feature of rustc_abi additive

[HKalbasi]

The goal here is to make rust-analyzer able to build with the rustc_private versions of the rustc crates it depends on. See #116847

[rustbot]

r? @petrochenkov

(rustbot has picked a reviewer for you, use r? to override)

[rustbot]

Some changes occurred to MIR optimizations

cc @rust-lang/wg-mir-opt

[fmease]

I just made an arbitrary field private in order to prevent the type from being constructible directly using fields. I'd appreciate suggestions to improve the code.

It's not very obvious / discoverable that this is the reason why repr was made private. At the very least, a comment on the field would help. However, I would recommend adding #[non_exhaustive] & a comment to ReprOptions or a marker field similar to AliasTy._use_alias_ty_new_instead.

[Noratrieb]

an alternative to this would also be to just remove the feature and make the seed a u64 instead. that doesn't really make the code much worse and should fit rust-analyzers needs

[HKalbasi]

I went this way. The feature couldn't be removed as it will enable the rand dependency which r-a had some problem with it IIRC. So now we are in a non ideal state were the field_shuffle_seed exists but has no effect without the randomize feature.

[Noratrieb]

r=me after updating the PR description

[HKalbasi]

@bors r=nilstrieb

[bors]

📌 Commit c8a25ed has been approved by nilstrieb

It is now in the queue for this repository.

[Noratrieb]

@bors r=Nilstrieb :)

[bors]

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

• There's another pull request that is currently being tested, blocking this pull request: [do not merge] CI experiments #112049

[bors]

📌 Commit c8a25ed has been approved by Nilstrieb

It is now in the queue for this repository.

[bors]

⌛ Testing commit c8a25ed with merge aea82b2...

[bors]

☀️ Test successful - checks-actions
Approved by: Nilstrieb
Pushing aea82b2 to master...

[rust-timer]

Finished benchmarking commit (aea82b2): comparison URL.

Overall result: no relevant changes - no action needed

@rustbot label: -perf-regression

Instruction count

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

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)	1.3%	[1.3%, 1.3%]	1
Regressions ❌ (secondary)	3.2%	[3.2%, 3.2%]	1
Improvements ✅ (primary)	-1.2%	[-1.9%, -0.4%]	2
Improvements ✅ (secondary)	-2.4%	[-3.7%, -1.0%]	2
All ❌✅ (primary)	-0.4%	[-1.9%, 1.3%]	3

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)	0.5%	[0.5%, 0.5%]	2
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-0.7%	[-0.7%, -0.6%]	2
Improvements ✅ (secondary)	-1.8%	[-1.8%, -1.8%]	1
All ❌✅ (primary)	-0.1%	[-0.7%, 0.5%]	4

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)	0.0%	[0.0%, 0.1%]	59
Regressions ❌ (secondary)	0.1%	[0.0%, 0.1%]	35
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	0.0%	[0.0%, 0.1%]	59

Bootstrap: 662.932s -> 663.447s (0.08%)
Artifact size: 308.95 MiB -> 308.96 MiB (0.00%)

[hlopko]

Thanks @HKalbasi so much for working on this!!