2023-06-21 Triage Log

Fairly positive week with very few regressions and some decent (albeit incremental) improvements. The most widespread gains came from some standard library improvements which are not a particularly common cause of large improvements in compiler performance. Overall a large chunk of the perf test suite showed an average of 0.6% improvement.

Triage done by @rylev. Revision range: 4bd4e2e..b9d608c

Summary:

(instructions:u)	mean	range	count
Regressions ❌ (primary)	1.4%	[0.5%, 3.6%]	16
Regressions ❌ (secondary)	1.0%	[0.2%, 3.5%]	29
Improvements ✅ (primary)	-0.8%	[-7.8%, -0.2%]	157
Improvements ✅ (secondary)	-2.9%	[-48.3%, -0.3%]	84
All ❌✅ (primary)	-0.6%	[-7.8%, 3.6%]	173

3 Regressions, 7 Improvements, 4 Mixed; 3 of them in rollups 51 artifact comparisons made in total

Regressions

Rollup of 7 pull requests #112625 (Comparison Link)

(instructions:u)	mean	range	count
Regressions ❌ (primary)	0.3%	[0.2%, 0.4%]	4
Regressions ❌ (secondary)	0.2%	[0.2%, 0.2%]	1
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	0.3%	[0.2%, 0.4%]	4

Seems like this was a blip in the bitmaps test which corrected itself in the next run.

Add AliasKind::Weak for type aliases. #108860 (Comparison Link)

(instructions:u)	mean	range	count
Regressions ❌ (primary)	0.3%	[0.2%, 0.5%]	8
Regressions ❌ (secondary)	0.4%	[0.2%, 0.8%]	7
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	0.3%	[0.2%, 0.5%]	8

Poked author and reviewer for thoughts on small regressions

CI: merge mingw test CI jobs #112645 (Comparison Link)

(instructions:u)	mean	range	count
Regressions ❌ (primary)	0.8%	[0.7%, 0.8%]	3
Regressions ❌ (secondary)	0.7%	[0.6%, 0.7%]	2
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	0.8%	[0.7%, 0.8%]	3

Noise as Linux was impacted

Improvements

Introduce a minimum CGU size in non-incremental builds. #112448 (Comparison Link)

(instructions:u)	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-2.1%	[-7.8%, -0.3%]	19
Improvements ✅ (secondary)	-8.5%	[-45.9%, -0.4%]	22
All ❌✅ (primary)	-2.1%	[-7.8%, -0.3%]	19

Don't drain-on-drop in DrainFilter impls of various collections. #104455 (Comparison Link)

(instructions:u)	mean	range	count
Regressions ❌ (primary)	0.5%	[0.2%, 1.2%]	3
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-0.5%	[-1.8%, -0.1%]	35
Improvements ✅ (secondary)	-0.5%	[-1.1%, -0.2%]	14
All ❌✅ (primary)	-0.4%	[-1.8%, 1.2%]	38

optimize slice::Iter::fold #106343 (Comparison Link)

(instructions:u)	mean	range	count
Regressions ❌ (primary)	0.7%	[0.4%, 1.1%]	5
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-0.2%	[-0.4%, -0.2%]	87
Improvements ✅ (secondary)	-0.3%	[-1.2%, -0.1%]	24
All ❌✅ (primary)	-0.2%	[-0.4%, 1.1%]	92

Rollup of 7 pull requests #112716 (Comparison Link)

(instructions:u)	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-0.4%	[-0.4%, -0.4%]	2
Improvements ✅ (secondary)	-0.4%	[-0.4%, -0.2%]	4
All ❌✅ (primary)	-0.4%	[-0.4%, -0.4%]	2

Rollup of 5 pull requests #112755 (Comparison Link)

(instructions:u)	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-0.8%	[-1.8%, -0.2%]	16
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-0.8%	[-1.8%, -0.2%]	16

Rollup of 6 pull requests #112774 (Comparison Link)

(instructions:u)	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-0.8%	[-0.8%, -0.7%]	3
Improvements ✅ (secondary)	-0.7%	[-0.7%, -0.7%]	2
All ❌✅ (primary)	-0.8%	[-0.8%, -0.7%]	3

new solver proof tree generation #112351 (Comparison Link)

(instructions:u)	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-0.5%	[-0.8%, -0.2%]	9
Improvements ✅ (secondary)	-1.2%	[-1.6%, -0.5%]	8
All ❌✅ (primary)	-0.5%	[-0.8%, -0.2%]	9

Mixed

Add support for targets without unwinding in mir-opt, and improve --bless for it #112418 (Comparison Link)

(instructions:u)	mean	range	count
Regressions ❌ (primary)	0.3%	[0.2%, 0.3%]	3
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-1.3%	[-1.3%, -1.3%]	1
All ❌✅ (primary)	0.3%	[0.2%, 0.3%]	3

Noise returning to a steady state

Ignore the always part of #[inline(always)] in MIR inlining #112294 (Comparison Link)

(instructions:u)	mean	range	count
Regressions ❌ (primary)	1.7%	[1.2%, 2.5%]	8
Regressions ❌ (secondary)	0.6%	[0.2%, 1.4%]	7
Improvements ✅ (primary)	-0.4%	[-3.0%, -0.2%]	61
Improvements ✅ (secondary)	-0.4%	[-1.0%, -0.2%]	18
All ❌✅ (primary)	-0.1%	[-3.0%, 2.5%]	69

From the author: “There are some notable regressions, and I think we might be able to claw those back by turning #[inline(always)] into a stronger hint. But I think this PR stands decently on its own as a tidy simplification.”

Remove box_free lang item #100036 (Comparison Link)

(instructions:u)	mean	range	count
Regressions ❌ (primary)	1.2%	[1.0%, 1.7%]	4
Regressions ❌ (secondary)	1.7%	[1.1%, 3.4%]	5
Improvements ✅ (primary)	-0.6%	[-0.6%, -0.6%]	2
Improvements ✅ (secondary)	-2.5%	[-4.4%, -1.4%]	4
All ❌✅ (primary)	0.6%	[-0.6%, 1.7%]	6

Waiting on the author and reviewer to give their thoughts.

make mir dataflow graphviz dumps opt-in #112617 (Comparison Link)

(instructions:u)	mean	range	count
Regressions ❌ (primary)	0.6%	[0.6%, 0.7%]	3
Regressions ❌ (secondary)	0.6%	[0.6%, 0.6%]	1
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-0.5%	[-0.6%, -0.4%]	6
All ❌✅ (primary)	0.6%	[0.6%, 0.7%]	3

Regressions are noise