bootstrap: build std sans leaf frame pointers #141800

workingjubilee · 2025-05-30T22:57:31Z

Sometimes leaf frame-pointers can impact LLVM inlining choices, and that can be a real problem for things like mul_add.

Sometimes leaf frame-pointers can impact LLVM inlining choices, and that can be a real problem for things like `mul_add`.

workingjubilee · 2025-05-31T00:30:09Z

@bors2 try @rust-timer queue

rust-bors · 2025-05-31T00:30:13Z

⌛ Trying commit 2d5d55a with merge 53f6b1b…

…ers, r=<try> bootstrap: build std with less frame pointers Sometimes leaf frame-pointers can impact LLVM inlining choices, and that can be a real problem for things like `mul_add`. r? `@ghost`

rust-bors · 2025-05-31T02:52:14Z

☀️ Try build successful

CI ✅
Build commit: 53f6b1b (53f6b1bd2116ea8d2f8cbd88780fa56984c18c0e)

rust-timer · 2025-05-31T21:48:49Z

Finished benchmarking commit (53f6b1b): comparison URL.

Overall result: ✅ improvements - no action needed

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.

@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
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-0.3%	[-0.4%, -0.2%]	3
All ❌✅ (primary)	-	-	0

Max RSS (memory usage)

Results (secondary -0.4%)

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
Regressions ❌ (secondary)	0.5%	[0.5%, 0.5%]	2
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-0.7%	[-0.9%, -0.4%]	7
All ❌✅ (primary)	-	-	0

Cycles

Results (secondary -0.6%)

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
Regressions ❌ (secondary)	1.0%	[0.7%, 1.2%]	2
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-0.8%	[-2.4%, -0.4%]	13
All ❌✅ (primary)	-	-	0

Binary size

Results (primary -0.3%, secondary -0.6%)

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
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-0.3%	[-0.6%, -0.1%]	11
Improvements ✅ (secondary)	-0.6%	[-0.6%, -0.4%]	38
All ❌✅ (primary)	-0.3%	[-0.6%, -0.1%]	11

Bootstrap: 778.921s -> 776.952s (-0.25%)
Artifact size: 370.25 MiB -> 370.14 MiB (-0.03%)

workingjubilee · 2025-05-31T22:02:55Z

Performance is almost completely invisible on a large, noisy program like rustc. But the improvements are more notable in binary size where they are modest but consistent. Not sure how to interpret the wall-time regressions in large-workspace, though.

saethlin · 2025-05-31T22:07:59Z

Not sure how to interpret the wall-time regressions in large-workspace, though.

Get a new commit hash and try again. It's probably instability.

saethlin · 2025-05-31T22:09:21Z

Oh. Most of the large-workspace significance thresholds are still the default 0.20%. The significance estimations just aren't calibrated yet.

workingjubilee · 2025-06-01T00:49:22Z

I'm gonna assume that means I shouldn't worry too much about it.

r? compiler

saethlin · 2025-06-01T01:26:29Z

Yeah, there's just no information to be gleaned from that benchmark. Yet.

I don't know what the right reviewer group is for "how we build the precompiled sysroot" but I've been involved in this discussion a fair bit and support this change.

r? saethlin
@bors r+

bors · 2025-06-01T01:26:32Z

📌 Commit 2d5d55a has been approved by saethlin

It is now in the queue for this repository.

bors · 2025-06-06T21:22:03Z

⌛ Testing commit 2d5d55a with merge 868bf2d...

bors · 2025-06-07T00:19:52Z

☀️ Test successful - checks-actions
Approved by: saethlin
Pushing 868bf2d to master...

github-actions · 2025-06-07T00:22:54Z

What is this?

This is an experimental post-merge analysis report that shows differences in test outcomes between the merged PR and its parent PR.

Comparing 44f415c (parent) -> 868bf2d (this PR)

Test differences

No test diffs found

Test dashboard

Run

cargo run --manifest-path src/ci/citool/Cargo.toml -- \
    test-dashboard 868bf2da31d1c6ead7016d5ee52eacfd2e2fd158 --output-dir test-dashboard

And then open test-dashboard/index.html in your browser to see an overview of all executed tests.

Job duration changes

x86_64-apple-1: 9697.9s -> 6534.4s (-32.6%)
x86_64-apple-2: 5212.7s -> 3655.8s (-29.9%)
dist-aarch64-apple: 4680.0s -> 5426.0s (15.9%)
mingw-check-1: 1898.4s -> 1614.7s (-14.9%)
dist-x86_64-apple: 9637.5s -> 8274.4s (-14.1%)
x86_64-rust-for-linux: 2947.4s -> 2531.7s (-14.1%)
i686-gnu-2: 6160.4s -> 5462.0s (-11.3%)
x86_64-gnu-llvm-19-1: 3561.9s -> 3224.4s (-9.5%)
x86_64-gnu-aux: 6515.8s -> 5918.6s (-9.2%)
x86_64-gnu-llvm-20-2: 5954.7s -> 5434.4s (-8.7%)

How to interpret the job duration changes?

Job durations can vary a lot, based on the actual runner instance
that executed the job, system noise, invalidated caches, etc. The table above is provided
mostly for t-infra members, for simpler debugging of potential CI slow-downs.

rust-timer · 2025-06-07T01:49:22Z

Finished benchmarking commit (868bf2d): comparison URL.

Overall result: ✅ improvements - no action needed

@rustbot label: -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
Improvements ✅ (primary)	-0.2%	[-0.2%, -0.2%]	1
Improvements ✅ (secondary)	-0.2%	[-0.2%, -0.2%]	3
All ❌✅ (primary)	-0.2%	[-0.2%, -0.2%]	1

Max RSS (memory usage)

Results (primary 4.2%, secondary -4.6%)

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)	4.2%	[4.2%, 4.2%]	1
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-4.6%	[-4.6%, -4.6%]	1
All ❌✅ (primary)	4.2%	[4.2%, 4.2%]	1

Cycles

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

Binary size

Results (primary -0.2%, secondary -0.5%)

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
Regressions ❌ (secondary)	0.0%	[0.0%, 0.0%]	1
Improvements ✅ (primary)	-0.2%	[-0.6%, -0.1%]	14
Improvements ✅ (secondary)	-0.5%	[-0.6%, -0.3%]	41
All ❌✅ (primary)	-0.2%	[-0.6%, -0.1%]	14

Bootstrap: 752.053s -> 751.358s (-0.09%)
Artifact size: 372.47 MiB -> 372.51 MiB (0.01%)

bootstrap: build std with less frame pointers

2d5d55a

Sometimes leaf frame-pointers can impact LLVM inlining choices, and that can be a real problem for things like `mul_add`.

rustbot added the T-bootstrap label May 30, 2025

This comment has been minimized.

Sign in to view

rustbot added the S-waiting-on-perf label May 31, 2025

This comment has been minimized.

Sign in to view

workingjubilee changed the title ~~bootstrap: build std with less frame pointers~~ bootstrap: build std sans leaf frame pointers May 31, 2025

rustbot removed the S-waiting-on-perf label May 31, 2025

workingjubilee marked this pull request as ready for review May 31, 2025 22:03

rustbot added the S-waiting-on-review label May 31, 2025

rustbot added the T-compiler label Jun 1, 2025

rustbot assigned jieyouxu Jun 1, 2025

rustbot assigned saethlin and unassigned jieyouxu Jun 1, 2025

bors added S-waiting-on-bors and removed S-waiting-on-review labels Jun 1, 2025

bors added the merged-by-bors label Jun 7, 2025

bors merged commit 868bf2d into rust-lang:master Jun 7, 2025
10 checks passed

rustbot added this to the 1.89.0 milestone Jun 7, 2025

workingjubilee deleted the build-std-with-less-leaf-pointers branch June 7, 2025 01:38

bootstrap: build std sans leaf frame pointers #141800

bootstrap: build std sans leaf frame pointers #141800

Uh oh!

Conversation

workingjubilee commented May 30, 2025 • edited by saethlin Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

workingjubilee commented May 31, 2025

Uh oh!

This comment has been minimized.

rust-bors bot commented May 31, 2025

Uh oh!

rust-bors bot commented May 31, 2025

Uh oh!

This comment has been minimized.

rust-timer commented May 31, 2025

Overall result: ✅ improvements - no action needed

Uh oh!

workingjubilee commented May 31, 2025

Uh oh!

saethlin commented May 31, 2025

Uh oh!

saethlin commented May 31, 2025

Uh oh!

workingjubilee commented Jun 1, 2025

Uh oh!

saethlin commented Jun 1, 2025

Uh oh!

bors commented Jun 1, 2025

Uh oh!

bors commented Jun 6, 2025

Uh oh!

bors commented Jun 7, 2025

Uh oh!

Uh oh!

github-actions bot commented Jun 7, 2025

Test differences

Job duration changes

Uh oh!

rust-timer commented Jun 7, 2025

Overall result: ✅ improvements - no action needed

Uh oh!

Uh oh!

workingjubilee commented May 30, 2025 •

edited by saethlin

Loading