[Perf experiment] Inline hints on some small, hot functions

[FractalFir]

During some local profiling, I identified a few tiny hot functions which did not get inlined.

From some local perf runs, it looks like adding inline hints to them is beneficial.

I have checked this change both with and without LTO, and I still see a perf difference between the version with and without inline hints. So, this may be a good change.

However, since I can't run the compiler PGO setup locally, I'll need a perf run to see if this improvement will be seen in the CI build of rustc.

There is a chance this optimization is already done by PGO(in which case this PR is less useful), but I still want to see if this can have any real effect.

[rustbot]

r? @spastorino

rustbot has assigned @spastorino.
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

[FractalFir]

Perf run to see if there is any merit to this change:

@bors try @rust-timer queue

[bors]

⌛ Trying commit aa1d580 with merge dd153da...

[bors]

☀️ Try build successful - checks-actions
Build commit: dd153da (dd153da1190de1ad17748851f0d9ab4d1e17edd5)

[rust-timer]

Finished benchmarking commit (dd153da): comparison URL.

Overall result: ❌✅ regressions and improvements - please read the text below

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 the most reliable metric that we have; it was used to determine the overall result at the top of this comment. However, even this metric can sometimes exhibit noise.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	0.5%	[0.5%, 0.5%]	1
Improvements ✅ (primary)	-1.3%	[-1.3%, -1.3%]	1
Improvements ✅ (secondary)	-0.7%	[-0.7%, -0.7%]	1
All ❌✅ (primary)	-1.3%	[-1.3%, -1.3%]	1

Max RSS (memory usage)

Results (primary 0.3%, secondary 0.7%)

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)	0.9%	[0.4%, 3.4%]	48
Regressions ❌ (secondary)	0.8%	[0.4%, 3.4%]	76
Improvements ✅ (primary)	-0.9%	[-2.3%, -0.4%]	23
Improvements ✅ (secondary)	-1.2%	[-2.1%, -0.4%]	7
All ❌✅ (primary)	0.3%	[-2.3%, 3.4%]	71

Cycles

Results (primary -0.2%, secondary -0.2%)

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)	1.1%	[0.4%, 4.2%]	29
Regressions ❌ (secondary)	0.9%	[0.4%, 3.3%]	44
Improvements ✅ (primary)	-1.2%	[-4.9%, -0.4%]	36
Improvements ✅ (secondary)	-0.9%	[-4.5%, -0.4%]	68
All ❌✅ (primary)	-0.2%	[-4.9%, 4.2%]	65

Binary size

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

Bootstrap: 770.447s -> 770.256s (-0.02%)
Artifact size: 359.63 MiB -> 359.63 MiB (0.00%)

[saethlin]

I have checked this change both with and without LTO,

Did you benchmark a stage2 build? LTO is ignored for stage1 builds.

[FractalFir]

I have checked this change both with and without LTO,

Did you benchmark a stage2 build? LTO is ignored for stage1 builds.

I just used ./x perf benchmark. I think it was a stage 2 build.