EXPERIMENT: Use the force-inline pass for more stuff #144483

scottmcm · 2025-07-26T04:40:48Z

We're already paying for the extra look at stuff, what if we add a second attribute for things that we always want to inline, but are also fine with generic calls not getting inlined? Can we save a bunch of code and functions calls in derived Ords and such? Is it worth the extra overhead of normalizing more calls?

r? ghost

EXPERIMENT: Use the force-inline pass for more stuff

scottmcm · 2025-07-26T09:04:01Z

@bors try @rust-timer queue

EXPERIMENT: Use the force-inline pass for more stuff

rust-bors · 2025-07-26T09:04:05Z

⌛ Trying commit e9fcbb2 with merge 5a3178f…

To cancel the try build, run the command @bors try cancel.

rust-bors · 2025-07-26T11:19:37Z

☀️ Try build successful (CI)
Build commit: 5a3178f (5a3178fa03ad01debfab17b1253662bc45c95eb9, parent: f32b23204a0efe2fe8383ed4be1a30b56c1bbf94)

rust-timer · 2025-07-26T12:59:40Z

Finished benchmarking commit (5a3178f): comparison URL.

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

Benchmarking this pull request means it may be perf-sensitive – we'll automatically label it not fit for rolling up. You can override this, but we strongly advise not to, due to possible changes in compiler perf.

Next Steps: If you can justify the regressions found in this try perf run, please do so in sufficient writing along with @rustbot label: +perf-regression-triaged. If not, please fix the regressions and do another perf run. If its results are neutral or positive, the label will be automatically removed.

@bors rollup=never
@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)	0.9%	[0.1%, 3.2%]	93
Regressions ❌ (secondary)	1.0%	[0.1%, 5.2%]	73
Improvements ✅ (primary)	-0.8%	[-2.6%, -0.1%]	5
Improvements ✅ (secondary)	-0.1%	[-0.2%, -0.1%]	2
All ❌✅ (primary)	0.8%	[-2.6%, 3.2%]	98

Max RSS (memory usage)

Results (primary 1.8%, 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)	2.3%	[0.6%, 6.2%]	23
Regressions ❌ (secondary)	3.8%	[1.1%, 6.5%]	2
Improvements ✅ (primary)	-3.6%	[-6.3%, -0.9%]	2
Improvements ✅ (secondary)	-1.5%	[-2.2%, -0.7%]	4
All ❌✅ (primary)	1.8%	[-6.3%, 6.2%]	25

Cycles

Results (primary 2.3%, secondary 2.8%)

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

	mean	range	count
Regressions ❌ (primary)	2.7%	[2.2%, 3.7%]	12
Regressions ❌ (secondary)	2.8%	[1.5%, 4.4%]	12
Improvements ✅ (primary)	-1.9%	[-1.9%, -1.9%]	1
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	2.3%	[-1.9%, 3.7%]	13

Binary size

Results (primary 1.3%, secondary 0.9%)

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

	mean	range	count
Regressions ❌ (primary)	1.4%	[0.0%, 8.5%]	96
Regressions ❌ (secondary)	1.0%	[0.1%, 2.4%]	21
Improvements ✅ (primary)	-0.2%	[-0.2%, -0.0%]	4
Improvements ✅ (secondary)	-0.1%	[-0.3%, -0.0%]	4
All ❌✅ (primary)	1.3%	[-0.2%, 8.5%]	100

Bootstrap: 467.999s -> 468.02s (0.00%)
Artifact size: 376.62 MiB -> 376.79 MiB (0.05%)

bors · 2025-07-26T18:14:24Z

☔ The latest upstream changes (presumably #144502) made this pull request unmergeable. Please resolve the merge conflicts.

scottmcm · 2025-07-26T18:24:04Z

Hmm, yeah, way too many new resolve calls to work. Ah well, interesting to know.

scottmcm · 2025-08-04T00:29:32Z

I was inspired by https://discord.com/channels/273534239310479360/957720175619215380/1401622766536364124 to see if I can do a more limited version of this that would still work.

…ther than `always`) Obviously there's way more that could do this, but I don't to do *all* of them at once.

scottmcm · 2025-08-04T02:55:41Z

@bors try @rust-timer queue

rust-bors · 2025-08-04T02:55:45Z

⌛ Trying commit a6adf2d with merge f4b36c8…

To cancel the try build, run the command @bors try cancel.

EXPERIMENT: Use the force-inline pass for more stuff

rust-log-analyzer · 2025-08-04T03:27:21Z

The job aarch64-gnu-llvm-19-1 failed! Check out the build log: (web) (plain enhanced) (plain)

Click to see the possible cause of the failure (guessed by this bot)

failures:

---- [ui] tests/ui-fulldeps/rustc_public/crate-info.rs stdout ----

error: test did not exit with success! code=Some(101) so test would pass with `run-fail`
status: exit status: 101
command: cd "/checkout/obj/build/aarch64-unknown-linux-gnu/test/ui-fulldeps/rustc_public/crate-info" && RUSTC="/checkout/obj/build/aarch64-unknown-linux-gnu/stage2/bin/rustc" RUST_TEST_THREADS="4" "/checkout/obj/build/aarch64-unknown-linux-gnu/test/ui-fulldeps/rustc_public/crate-info/a"
stdout: none
--- stderr -------------------------------

Add some pre-codegen MIR tests for debug mode No functional changes; just some tests. I made these for rust-lang#144483, but that's going in a different direction, so I wanted to propose we just add them to help see the impact of other related changes in the future. r? mir

rust-bors · 2025-08-04T05:11:21Z

☀️ Try build successful (CI)
Build commit: f4b36c8 (f4b36c8b56f988af2e104136c467ff4a28a21216, parent: 383b9c447b61641e1f1a3850253944a897a60827)

rust-timer · 2025-08-04T07:27:56Z

Finished benchmarking commit (f4b36c8): comparison URL.

Overall result: ❌ regressions - please read the text below

Benchmarking this pull request means it may be perf-sensitive – we'll automatically label it not fit for rolling up. You can override this, but we strongly advise not to, due to possible changes in compiler perf.

Next Steps: If you can justify the regressions found in this try perf run, please do so in sufficient writing along with @rustbot label: +perf-regression-triaged. If not, please fix the regressions and do another perf run. If its results are neutral or positive, the label will be automatically removed.

@bors rollup=never
@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)	0.8%	[0.1%, 1.0%]	7
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-0.0%	[-0.0%, -0.0%]	1
All ❌✅ (primary)	-	-	0

Max RSS (memory usage)

Results (primary -0.7%, secondary -2.1%)

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

	mean	range	count
Regressions ❌ (primary)	1.7%	[0.9%, 2.4%]	4
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-5.4%	[-7.4%, -3.4%]	2
Improvements ✅ (secondary)	-2.1%	[-3.1%, -0.8%]	4
All ❌✅ (primary)	-0.7%	[-7.4%, 2.4%]	6

Cycles

Results (secondary -4.7%)

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)	-	-	0
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-4.7%	[-4.7%, -4.7%]	1
All ❌✅ (primary)	-	-	0

Binary size

Results (primary -0.0%)

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

	mean	range	count
Regressions ❌ (primary)	0.0%	[0.0%, 0.1%]	12
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-0.1%	[-0.1%, -0.0%]	7
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-0.0%	[-0.1%, 0.1%]	19

Bootstrap: 468.859s -> 468.747s (-0.02%)
Artifact size: 376.98 MiB -> 377.03 MiB (0.01%)

Rollup merge of #144875 - scottmcm:more-mir-tests, r=cjgillot Add some pre-codegen MIR tests for debug mode No functional changes; just some tests. I made these for #144483, but that's going in a different direction, so I wanted to propose we just add them to help see the impact of other related changes in the future. r? mir

This comment was marked as outdated.

Sign in to view

rust-bors bot added a commit that referenced this pull request Jul 26, 2025

Auto merge of #144483 - scottmcm:early-inline, r=<try>

e41d15b

EXPERIMENT: Use the force-inline pass for more stuff

rustbot added the S-waiting-on-perf Status: Waiting on a perf run to be completed. label Jul 26, 2025

This comment has been minimized.

Sign in to view

This comment was marked as outdated.

Sign in to view

scottmcm force-pushed the early-inline branch from 35b935a to e9fcbb2 Compare July 26, 2025 07:03

This comment has been minimized.

Sign in to view

rust-bors bot added a commit that referenced this pull request Jul 26, 2025

Auto merge of #144483 - scottmcm:early-inline, r=<try>

5a3178f

EXPERIMENT: Use the force-inline pass for more stuff

This comment has been minimized.

Sign in to view

rustbot added perf-regression Performance regression. and removed S-waiting-on-perf Status: Waiting on a perf run to be completed. labels Jul 26, 2025

scottmcm closed this Jul 26, 2025

rustbot removed the S-waiting-on-author Status: This is awaiting some action (such as code changes or more information) from the author. label Jul 26, 2025

scottmcm reopened this Aug 4, 2025

rustbot added the S-waiting-on-author Status: This is awaiting some action (such as code changes or more information) from the author. label Aug 4, 2025

scottmcm mentioned this pull request Aug 4, 2025

Add some pre-codegen MIR tests for debug mode #144875

Merged

scottmcm force-pushed the early-inline branch from e9fcbb2 to 4e11683 Compare August 4, 2025 00:48

scottmcm added 3 commits August 3, 2025 19:21

Add a MIR test for integer methods in debug

7626d22

Add #[rustc_early_inline]

5a244aa

Make cast_(un)signed and wrapping_add be rustc_early_inline (ra…

a6adf2d

…ther than `always`) Obviously there's way more that could do this, but I don't to do *all* of them at once.

scottmcm force-pushed the early-inline branch from 4e11683 to a6adf2d Compare August 4, 2025 02:45

This comment has been minimized.

Sign in to view

rust-bors bot added a commit that referenced this pull request Aug 4, 2025

Auto merge of #144483 - scottmcm:early-inline, r=<try>

f4b36c8

EXPERIMENT: Use the force-inline pass for more stuff

rustbot added the S-waiting-on-perf Status: Waiting on a perf run to be completed. label Aug 4, 2025

This comment has been minimized.

Sign in to view

rustbot removed the S-waiting-on-perf Status: Waiting on a perf run to be completed. label Aug 4, 2025

EXPERIMENT: Use the force-inline pass for more stuff #144483

Are you sure you want to change the base?

EXPERIMENT: Use the force-inline pass for more stuff #144483

Uh oh!

Conversation

scottmcm commented Jul 26, 2025

Uh oh!

This comment was marked as outdated.

This comment was marked as outdated.

This comment was marked as outdated.

This comment has been minimized.

This comment has been minimized.

This comment was marked as outdated.

scottmcm commented Jul 26, 2025

Uh oh!

This comment has been minimized.

rust-bors bot commented Jul 26, 2025

Uh oh!

rust-bors bot commented Jul 26, 2025

Uh oh!

This comment has been minimized.

rust-timer commented Jul 26, 2025

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

Uh oh!

bors commented Jul 26, 2025

Uh oh!

scottmcm commented Jul 26, 2025

Uh oh!

scottmcm commented Aug 4, 2025

Uh oh!

scottmcm commented Aug 4, 2025

Uh oh!

This comment has been minimized.

rust-bors bot commented Aug 4, 2025

Uh oh!

rust-log-analyzer commented Aug 4, 2025

Uh oh!

rust-bors bot commented Aug 4, 2025

Uh oh!

This comment has been minimized.

rust-timer commented Aug 4, 2025

Overall result: ❌ regressions - please read the text below

Uh oh!

Uh oh!