Use MirPatch in simplify_branches.

[cjgillot]

This allows to avoid clearing the CFG cache if we don't perform any change.

r? @ghost for perf

[cjgillot]

@bors try @rust-timer queue

[rust-bors]

☀️ Try build successful (CI)
Build commit: 0e126a7 (0e126a774c708007b8130681b246a072de097971, parent: c8905eaa66e0c35a33626e974b9ce6955c739b5b)

[rust-timer]

Finished benchmarking commit (0e126a7): comparison URL.

Overall result: ❌ regressions - no action needed

Benchmarking this pull request means it may be perf-sensitive – we'll automatically label it not fit for rolling up. You can override this, but we strongly advise not to, due to possible changes in compiler perf.

@bors rollup=never
@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)	0.0%	[0.0%, 0.0%]	1
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-	-	0

Max RSS (memory usage)

Results (primary 0.7%, secondary -0.8%)

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

	mean	range	count
Regressions ❌ (primary)	2.1%	[1.1%, 2.7%]	3
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-3.6%	[-3.6%, -3.6%]	1
Improvements ✅ (secondary)	-0.8%	[-0.8%, -0.8%]	1
All ❌✅ (primary)	0.7%	[-3.6%, 2.7%]	4

Cycles

Results (secondary 2.8%)

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)	2.8%	[2.5%, 3.4%]	3
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-	-	0

Binary size

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

Bootstrap: 470.874s -> 470.61s (-0.06%)
Artifact size: 387.66 MiB -> 387.68 MiB (0.01%)

[rustbot]

Some changes occurred to MIR optimizations

cc @rust-lang/wg-mir-opt

[cjgillot]

r? compiler