rustc_resolve: Further reduce mutability in resolver

[petrochenkov]

resolver.cm() now returns a conditionally mutable reference (RefOrMut) that cannot be mutated, and resolver.cm_mut() returns a conditionally mutable reference that can be mutated.

The "can be mutated" property is checked dynamically, so this can potentially panic in some obscure error reporting cases not covered by the test suite, but they should be easily fixable and will be caught sooner or later, but it's more useful to see which code path actually use resolver in mutable way rather than just take RefOrMut.

The second commit converts RefOrMut to an enum instead of using raw pointers.

r? @mu001999
cc @LorrensP-2158466

[petrochenkov]

I'll try changing RefOrMut to an enum to avoid unsafe code.
(I remember why it wasn't made an enum originally, but it should no longer be relevant.)
@rustbot author

[rustbot]

Reminder, once the PR becomes ready for a review, use @rustbot ready.

[mu001999]

I'm not sure how difficult this would be or whether it's feasible, but could we use different types for Ref and Mut instead of combining them in RefOrMut?

[petrochenkov]

I'm not sure how difficult this would be or whether it's feasible, but could we use different types for Ref and Mut instead of combining them in RefOrMut?

Not feasible, functions taking RefOrMut do that because they must work with both mutable and immutable versions, dynamically.

Or rather infeasible without duplicating a large number of resolver functions.
Perhaps the "duplication" could also be done by turning Finalize into a const generic argument, although I'm not sure and it would also be a large change.

[petrochenkov]

@bors try @rust-timer queue

[rust-bors]

☀️ Try build successful (CI)
Build commit: df44b1a (df44b1aa081b3f74f4d68fe6d1de048820cbb3ca)
Base parent: ce69831 (ce6983167791bf9418726264f8e5cc7abf73d69b)

[rust-timer]

Finished benchmarking commit (df44b1a): comparison URL.

Overall result: ❌ regressions - no action needed

Benchmarking means the PR may be perf-sensitive. Consider adding rollup=never if this change is not fit for rolling up.

@rustbot label: -S-waiting-on-perf -perf-regression

Instruction count

Our most reliable metric. Used to determine the overall result above. However, even this metric can be noisy.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	0.2%	[0.2%, 0.2%]	1
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-	-	0

Max RSS (memory usage)

Results (primary 3.7%, secondary 0.4%)

A less reliable metric. May be of interest, but not used to determine the overall result above.

	mean	range	count
Regressions ❌ (primary)	3.7%	[1.4%, 5.2%]	3
Regressions ❌ (secondary)	5.3%	[5.3%, 5.3%]	1
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-4.6%	[-4.6%, -4.6%]	1
All ❌✅ (primary)	3.7%	[1.4%, 5.2%]	3

Cycles

Results (secondary 2.4%)

A less reliable metric. May be of interest, but not used to determine the overall result above.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	2.4%	[2.3%, 2.5%]	2
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-	-	0

Binary size

This perf run didn't have relevant results for this metric.

Bootstrap: 490.002s -> 491.705s (0.35%)
Artifact size: 390.14 MiB -> 390.21 MiB (0.02%)

[rustbot]

This PR was rebased onto a different main commit. Here's a range-diff highlighting what actually changed.

Rebasing is a normal part of keeping PRs up to date, so no action is needed—this note is just to help reviewers.

[petrochenkov]

@rustbot ready

[mu001999]

@rustbot author

View changes since this review

[mu001999]

@bors r+ rollup

[rust-bors]

📌 Commit c9f7b96 has been approved by mu001999

It is now in the queue for this repository.

[jhpratt]

@rust-timer build 3bb5bfc

[rust-timer]

Finished benchmarking commit (3bb5bfc): comparison URL.

Overall result: ✅ improvements - no action needed

Benchmarking means the PR may be perf-sensitive. Consider adding rollup=never if this change is not fit for rolling up.

@rustbot label: -S-waiting-on-perf -perf-regression

Instruction count

Our most reliable metric. Used to determine the overall result above. However, even this metric can be noisy.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-0.3%	[-0.3%, -0.2%]	2
All ❌✅ (primary)	-	-	0

Max RSS (memory usage)

Results (primary -1.2%, secondary 0.1%)

A less reliable metric. May be of interest, but not used to determine the overall result above.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	3.3%	[3.3%, 3.3%]	1
Improvements ✅ (primary)	-1.2%	[-1.2%, -1.2%]	1
Improvements ✅ (secondary)	-0.5%	[-0.6%, -0.4%]	6
All ❌✅ (primary)	-1.2%	[-1.2%, -1.2%]	1

Cycles

Results (primary 0.2%, secondary -3.3%)

A less reliable metric. May be of interest, but not used to determine the overall result above.

	mean	range	count
Regressions ❌ (primary)	0.7%	[0.6%, 1.1%]	6
Regressions ❌ (secondary)	1.4%	[0.6%, 2.9%]	3
Improvements ✅ (primary)	-1.5%	[-2.1%, -1.0%]	2
Improvements ✅ (secondary)	-4.4%	[-16.8%, -0.4%]	13
All ❌✅ (primary)	0.2%	[-2.1%, 1.1%]	8

Binary size

This perf run didn't have relevant results for this metric.

Bootstrap: 491.471s -> 489.765s (-0.35%)
Artifact size: 390.58 MiB -> 390.47 MiB (-0.03%)