[wip] terminator perf experiments

[panstromek]

View all comments

[panstromek]

hoping for no impact here for now

@bors try @rust-timer queue

[rust-bors]

☀️ Try build successful (CI)
Build commit: c9696d1 (c9696d172403366b5d28ed4bee1020bea22674ce)
Base parent: 008fa22 (008fa22ce3f8d3c8dfaca2e6486043c2b21851eb)

[rust-timer]

Finished benchmarking commit (c9696d1): 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)	0.3%	[0.3%, 0.3%]	1
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-	-	0

Max RSS (memory usage)

Results (primary 4.3%, secondary -4.1%)

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

	mean	range	count
Regressions ❌ (primary)	4.3%	[4.3%, 4.3%]	1
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-4.1%	[-6.5%, -1.7%]	2
All ❌✅ (primary)	4.3%	[4.3%, 4.3%]	1

Cycles

Results (secondary -1.0%)

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)	4.1%	[4.1%, 4.1%]	1
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-6.0%	[-6.0%, -6.0%]	1
All ❌✅ (primary)	-	-	0

Binary size

Results (primary 0.1%, 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)	0.1%	[0.0%, 0.2%]	41
Regressions ❌ (secondary)	0.1%	[0.0%, 0.1%]	41
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	0.1%	[0.0%, 0.2%]	41

Bootstrap: 489.346s -> 488.243s (-0.23%)
Artifact size: 387.69 MiB -> 387.82 MiB (0.03%)

[panstromek]

Good, but looks like the metadata is slightly bigger. Probably because it encodes enum instead of just values & targets arrays. We could optimize that by encoding it as two arrays in all cases.

[panstromek]

@bors try @rust-timer queue

[rust-bors]

☀️ Try build successful (CI)
Build commit: a75e38e (a75e38eed3fb412c5725c75ad8ccb15cbc512bd2)
Base parent: fdbd8ee (fdbd8ee70d9b68d98be873e1095d74cf9441f0c8)

[rust-timer]

Finished benchmarking commit (a75e38e): 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 0.6%, secondary 4.4%)

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

	mean	range	count
Regressions ❌ (primary)	3.0%	[3.0%, 3.0%]	1
Regressions ❌ (secondary)	4.4%	[4.4%, 4.4%]	1
Improvements ✅ (primary)	-1.7%	[-1.7%, -1.7%]	1
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	0.6%	[-1.7%, 3.0%]	2

Cycles

Results (secondary 3.4%)

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)	3.4%	[3.0%, 3.8%]	3
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-	-	0

Binary size

Results (primary 0.1%, 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)	0.1%	[0.0%, 0.2%]	37
Regressions ❌ (secondary)	0.1%	[0.0%, 0.1%]	40
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	0.1%	[0.0%, 0.2%]	37

Bootstrap: 487.865s -> 487.9s (0.01%)
Artifact size: 387.69 MiB -> 387.68 MiB (-0.00%)

[panstromek]

This is the first one that's supposed to have some impact - one byte down from TerminatorKind (and consequently Terminator and BasicBlockData), but achieved with ThinVec for Call args. That is less shrink_to_fit calls and possible reallocs, but also more heap usage for non-empty args.

@bors try @rust-timer queue

[rust-bors]

☀️ Try build successful (CI)
Build commit: 1338b91 (1338b91d2abc3d1e9cca32ce60d67321e973062f)
Base parent: 9451e06 (9451e069c34e1d7db23a155dcf9c25d1bc127055)

[rust-timer]

Finished benchmarking commit (1338b91): comparison URL.

Overall result: ❌ regressions - please read:

Benchmarking means the PR may be perf-sensitive. It's automatically marked not fit for rolling up. Overriding is possible but disadvised: it risks changing compiler perf.

Next, please: If you can, justify the regressions found in this try perf run in writing along with @rustbot label: +perf-regression-triaged. If not, fix the regressions and do another perf run. Neutral or positive results will clear the label automatically.

@bors rollup=never rustc-perf
@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)	0.6%	[0.2%, 2.6%]	19
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-0.2%	[-0.2%, -0.2%]	1
All ❌✅ (primary)	-	-	0

Max RSS (memory usage)

Results (primary 0.7%, 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)	1.0%	[0.5%, 2.1%]	12
Regressions ❌ (secondary)	1.8%	[1.2%, 2.4%]	2
Improvements ✅ (primary)	-3.1%	[-3.1%, -3.1%]	1
Improvements ✅ (secondary)	-2.3%	[-2.3%, -2.3%]	1
All ❌✅ (primary)	0.7%	[-3.1%, 2.1%]	13

Cycles

Results (secondary 3.7%)

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)	3.7%	[2.2%, 5.2%]	2
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-	-	0

Binary size

Results (primary 0.1%, 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)	0.1%	[0.0%, 0.2%]	38
Regressions ❌ (secondary)	0.1%	[0.0%, 0.1%]	40
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	0.1%	[0.0%, 0.2%]	38

Bootstrap: 489.176s -> 487.69s (-0.30%)
Artifact size: 387.73 MiB -> 387.69 MiB (-0.01%)

[panstromek]

hm. I wonder what's up with the new solver that it stresses this code that much. That doesn't seem super related.

[panstromek]

new solver changes were caused by thin-vec update. Those were resolved in other PRs, I also got a better Decodable impl from review on other PR. Let's try to test the impact after all those changes. I also tweaked take_array to avoid reallocating thinvec to box on error (bug!) path (probably negligible, but maybe optimizes better?).

@bors try @rust-timer queue

[rust-bors]

☀️ Try build successful (CI)
Build commit: ec6ec96 (ec6ec96bab4d0725e243de12ac839706103c3552)
Base parent: fc0f51f (fc0f51f5ca44d4586b0e309181382608cd6a8441)

[rust-timer]

Finished benchmarking commit (ec6ec96): comparison URL.

Overall result: ❌✅ regressions and 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)	0.7%	[0.7%, 0.7%]	2
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-0.2%	[-0.3%, -0.2%]	2
All ❌✅ (primary)	-	-	0

Max RSS (memory usage)

Results (primary 0.9%, 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)	1.1%	[0.6%, 3.6%]	20
Regressions ❌ (secondary)	1.4%	[0.4%, 2.4%]	2
Improvements ✅ (primary)	-0.5%	[-0.6%, -0.5%]	4
Improvements ✅ (secondary)	-0.6%	[-1.0%, -0.5%]	6
All ❌✅ (primary)	0.9%	[-0.6%, 3.6%]	24

Cycles

Results (primary -0.7%, secondary 0.7%)

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

	mean	range	count
Regressions ❌ (primary)	0.5%	[0.5%, 0.5%]	1
Regressions ❌ (secondary)	1.7%	[0.5%, 4.0%]	8
Improvements ✅ (primary)	-0.9%	[-1.5%, -0.5%]	10
Improvements ✅ (secondary)	-0.7%	[-0.9%, -0.4%]	6
All ❌✅ (primary)	-0.7%	[-1.5%, 0.5%]	11

Binary size

Results (primary 0.1%, secondary 0.0%)

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

	mean	range	count
Regressions ❌ (primary)	0.1%	[0.0%, 0.2%]	48
Regressions ❌ (secondary)	0.1%	[0.0%, 0.1%]	41
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-0.1%	[-0.1%, -0.0%]	9
All ❌✅ (primary)	0.1%	[0.0%, 0.2%]	48

Bootstrap: 491.246s -> 488.936s (-0.47%)
Artifact size: 390.66 MiB -> 390.59 MiB (-0.02%)