Inline some hot new-solver functions

[nnethercote]

View all comments

rerunning_stalled_goal_may_make_progress is split into always-inline and never-inline variants, because it's biggish and has multiple call sites but only one is hot. The other are easy, being either very small or having a single call site.

r? @jdonszelmann

[nnethercote]

@bors try @rust-timer queue

[rust-bors]

☀️ Try build successful (CI)
Build commit: bdde1fb (bdde1fb32f7450db6fd5d778d69a117c761459ed)
Base parent: 0e72e32 (0e72e3266cdd2fe5a658aff454a778a88e5220a3)

[rust-timer]

Finished benchmarking commit (bdde1fb): comparison URL.

Overall result: no relevant changes - 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

This perf run didn't have relevant results for this metric.

Max RSS (memory usage)

Results (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)	-	-	0
Regressions ❌ (secondary)	0.5%	[0.5%, 0.5%]	1
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-0.6%	[-0.8%, -0.4%]	6
All ❌✅ (primary)	-	-	0

Cycles

Results (secondary -0.3%)

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)	0.7%	[0.5%, 1.0%]	5
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-0.7%	[-1.5%, -0.4%]	12
All ❌✅ (primary)	-	-	0

Binary size

This perf run didn't have relevant results for this metric.

Bootstrap: 454.011s -> 454.672s (0.15%)
Artifact size: 398.43 MiB -> 398.47 MiB (0.01%)

[nnethercote]

Interesting... no perf effect on CI, but I get big effects locally:

Presumably this is because CI has PGO and already inlines these functions based on profiling data.

I think it's worth merging this nonetheless, for two reasons:

• Not every platform has PGO, so the non-PGO platforms will benefit from this.
• People profiling locally won't be misled.

[rustbot]

r? @fee1-dead

rustbot has assigned @fee1-dead.
They will have a look at your PR within the next two weeks and either review your PR or reassign to another reviewer.

Use r? to explicitly pick a reviewer

Why was this reviewer chosen?

The reviewer was selected based on:

• Owners of files modified in this PR: compiler
• compiler expanded to 75 candidates
• Random selection from 17 candidates

[rustbot]

jdonszelmann is currently at their maximum review capacity.
They may take a while to respond.

[nnethercote]

I rebased over #158436 and somehow things got even better locally:

[nnethercote]

Let's do another perf run just to check it's still a no-op with PGO enabled.

@bors try @rust-timer queue

[rust-bors]

☀️ Try build successful (CI)
Build commit: 326d826 (326d826e4be272600a38d4309892f4631e3d5c6b)
Base parent: 3d6c19b (3d6c19bb9ab4798ecfb2ee943df01a811720fc27)

[rust-timer]

Finished benchmarking commit (326d826): comparison URL.

Overall result: no relevant changes - 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

This perf run didn't have relevant results for this metric.

Max RSS (memory usage)

Results (primary -3.0%, secondary -2.4%)

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

	mean	range	count
Regressions ❌ (primary)	2.8%	[2.8%, 2.8%]	1
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-5.8%	[-8.0%, -3.7%]	2
Improvements ✅ (secondary)	-2.4%	[-11.6%, -0.4%]	6
All ❌✅ (primary)	-3.0%	[-8.0%, 2.8%]	3

Cycles

Results (secondary 0.2%)

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)	0.7%	[0.4%, 1.1%]	14
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-0.7%	[-1.1%, -0.5%]	8
All ❌✅ (primary)	-	-	0

Binary size

This perf run didn't have relevant results for this metric.

Bootstrap: 455.92s -> 456.054s (0.03%)
Artifact size: 396.47 MiB -> 396.34 MiB (-0.03%)

[nnethercote]

Still neutral on CI, still worth landing IMO:

• Not every platform has PGO, so the non-PGO platforms will benefit from this.
• People profiling locally won't be misled.

[rustbot]

This PR was rebased onto a different main commit. Here's a range-diff highlighting what actually changed.

Rebasing is a normal part of keeping PRs up to date, so no action is needed—this note is just to help reviewers.

[nnethercote]

I rebased again. This is surprisingly conflict-prone for such a small PR. @jdonszelmann, if/when you get a chance... :)