Improve the CGU merging algorithm

[nnethercote]

r? @ghost

[nnethercote]

Measuring locally, this gave clearly better results -- more evenly sized CGUs -- but didn't have much effect on compile times, primarily because there are some big CGUs even before merging occurs, and they are the limiting factor. But let's try it on CI.

@bors try @rust-timer queue

[bors]

⌛ Trying commit a7f886e846acc849e05e0183061a5a584d53cbab with merge 520155aa1654c14240f3b2f4dcfab03a94968bfc...

[nnethercote]

Zulip discussion.

[bors]

☀️ Try build successful - checks-actions
Build commit: 520155aa1654c14240f3b2f4dcfab03a94968bfc (520155aa1654c14240f3b2f4dcfab03a94968bfc)

[rust-timer]

Finished benchmarking commit (520155aa1654c14240f3b2f4dcfab03a94968bfc): comparison URL.

Overall result: ❌✅ regressions and improvements - ACTION NEEDED

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 a highly reliable metric that was used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	0.9%	[0.2%, 2.6%]	12
Regressions ❌ (secondary)	0.7%	[0.2%, 1.6%]	14
Improvements ✅ (primary)	-0.9%	[-1.0%, -0.8%]	2
Improvements ✅ (secondary)	-1.4%	[-1.8%, -0.4%]	7
All ❌✅ (primary)	0.6%	[-1.0%, 2.6%]	14

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)	6.2%	[0.1%, 12.5%]	5
Regressions ❌ (secondary)	1.1%	[1.1%, 1.1%]	2
Improvements ✅ (primary)	-3.5%	[-5.5%, -2.1%]	5
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	1.4%	[-5.5%, 12.5%]	10

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)	2.7%	[2.7%, 2.7%]	1
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-2.6%	[-2.6%, -2.6%]	1
All ❌✅ (primary)	2.7%	[2.7%, 2.7%]	1

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)	0.4%	[0.0%, 1.6%]	24
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-0.1%	[-0.5%, -0.0%]	16
Improvements ✅ (secondary)	-0.6%	[-0.6%, -0.6%]	1
All ❌✅ (primary)	0.2%	[-0.5%, 1.6%]	40

Bootstrap: 641.625s -> 643.141s (0.24%)

[nnethercote]

Bad perf results all around: walltime, max-rss, and binary size are all worse on average. More work needed here. In the meantime I will open a new PR for the debug output improvements.