Use multiple returns in MIR if it saves a block; still have only one in LLVM

[scottmcm]

This is still an open question whether it's desired, per #72022 (comment), so opening as a draft.

Given that cranelift prefers multiple returns and we use optimized MIR for MIR inlining, it seems at least plausible that MIR shouldn't have the "one return" rule (the validator doesn't enforce it currently either) so we can have fewer total blocks, and then targets like LLVM can enforce it if they so choose (as this PR does).

r? ghost

[scottmcm]

@bors try @rust-timer queue

[bors]

⌛ Trying commit 2dd720c with merge be9306c...

[bors]

☀️ Try build successful - checks-actions
Build commit: be9306c (be9306cf3159eb27d96f4e1e2187e9dba4523c7d)

[rust-timer]

Finished benchmarking commit (be9306c): 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)	0.7%	[0.3%, 1.3%]	3
Regressions ❌ (secondary)	0.2%	[0.2%, 0.2%]	1
Improvements ✅ (primary)	-0.6%	[-0.7%, -0.6%]	2
Improvements ✅ (secondary)	-0.4%	[-0.4%, -0.4%]	1
All ❌✅ (primary)	0.1%	[-0.7%, 1.3%]	5

Max RSS (memory usage)

Results (primary -2.8%, secondary 2.3%)

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)	3.4%	[1.6%, 5.3%]	2
Regressions ❌ (secondary)	2.3%	[2.3%, 2.3%]	1
Improvements ✅ (primary)	-4.6%	[-7.8%, -2.4%]	7
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-2.8%	[-7.8%, 5.3%]	9

Cycles

Results (primary -0.7%, secondary -0.9%)

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)	-1.4%	[-2.3%, -1.0%]	3
Improvements ✅ (secondary)	-0.9%	[-0.9%, -0.9%]	1
All ❌✅ (primary)	-0.7%	[-2.3%, 1.5%]	4

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)	0.1%	[0.0%, 0.2%]	24
Regressions ❌ (secondary)	0.2%	[0.0%, 0.6%]	4
Improvements ✅ (primary)	-0.1%	[-1.1%, -0.0%]	45
Improvements ✅ (secondary)	-0.1%	[-0.3%, -0.0%]	14
All ❌✅ (primary)	-0.0%	[-1.1%, 0.2%]	69

Bootstrap: 765.289s -> 766.233s (0.12%)
Artifact size: 362.07 MiB -> 362.04 MiB (-0.01%)