Skip to content

Add perf for 2024-02-27 #1825

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 1 commit into from
Feb 27, 2024
Merged

Add perf for 2024-02-27 #1825

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
203 changes: 203 additions & 0 deletions triage/2024-02-27.md
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,203 @@
# 2024-02-27 Triage Log

A rare week where regressions out powered improvements to make the compiler roughly half a percent slower on average on nearly 100 benchmarks. Some regressions have fixes in the pipeline, but some remain elusive or were introduced to address correctness issues.

Triage done by **@rylev**.
Revision range: [5af21304..71ffdf7f](https://perf.rust-lang.org/?start=5af2130440c198afefbe5b8099342057cf272ef4&end=71ffdf7ff7ac6df5f9f64de7e780b8345797e8a0&absolute=false&stat=instructions%3Au)

**Summary**:

| (instructions:u) | mean | range | count |
|:----------------------------------:|:-----:|:--------------:|:-----:|
| Regressions ❌ <br /> (primary) | 1.0% | [0.2%, 4.4%] | 69 |
| Regressions ❌ <br /> (secondary) | 1.4% | [0.2%, 4.9%] | 66 |
| Improvements ✅ <br /> (primary) | -1.1% | [-3.3%, -0.2%] | 28 |
| Improvements ✅ <br /> (secondary) | -0.6% | [-1.5%, -0.2%] | 33 |
| All ❌✅ (primary) | 0.4% | [-3.3%, 4.4%] | 97 |


4 Regressions, 6 Improvements, 5 Mixed; 2 of them in rollups
58 artifact comparisons made in total

#### Regressions

Always evaluate free constants and statics, even if previous errors occurred [#121087](https://github.com/rust-lang/rust/pull/121087) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=5af2130440c198afefbe5b8099342057cf272ef4&end=cce6a6e22e715bd74455f2560a956ab920c3a914&stat=instructions:u)

| (instructions:u) | mean | range | count |
|:----------------------------------:|:----:|:------------:|:-----:|
| Regressions ❌ <br /> (primary) | 0.8% | [0.4%, 2.0%] | 4 |
| Regressions ❌ <br /> (secondary) | 1.0% | [0.3%, 5.2%] | 11 |
| Improvements ✅ <br /> (primary) | - | - | 0 |
| Improvements ✅ <br /> (secondary) | - | - | 0 |
| All ❌✅ (primary) | 0.8% | [0.4%, 2.0%] | 4 |
- Regressions in primary benchmarks are real and addressed in #121387
- Regressions in secondary benchmarks seem to be all noise.


Use intrinsics::debug_assertions in debug_assert_nounwind [#120863](https://github.com/rust-lang/rust/pull/120863) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=29f87ade9d78d233e85ef6ca2d6153d0d4fd38d6&end=2b43e75c98cc5ae32328c8b49657bcd882eb5e75&stat=instructions:u)

| (instructions:u) | mean | range | count |
|:----------------------------------:|:-----:|:--------------:|:-----:|
| Regressions ❌ <br /> (primary) | 0.8% | [0.3%, 1.7%] | 13 |
| Regressions ❌ <br /> (secondary) | - | - | 0 |
| Improvements ✅ <br /> (primary) | -0.4% | [-0.4%, -0.4%] | 1 |
| Improvements ✅ <br /> (secondary) | - | - | 0 |
| All ❌✅ (primary) | 0.7% | [-0.4%, 1.7%] | 14 |
- Pinged the author on what next steps are. Regression is in codegen which would be expected given the nature of the change, but it's unclear how one would address the issue.


wasm: Store rlib metadata in wasm object files [#120588](https://github.com/rust-lang/rust/pull/120588) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=0987e41d1ccbae7a8a55fdcde4af0ea76ebfe2d3&end=7168c13579a550f2c47f7eea22f5e226a436cd00&stat=instructions:u)

| (instructions:u) | mean | range | count |
|:----------------------------------:|:----:|:------------:|:-----:|
| Regressions ❌ <br /> (primary) | 3.1% | [3.0%, 3.3%] | 4 |
| Regressions ❌ <br /> (secondary) | 1.5% | [0.2%, 2.9%] | 33 |
| Improvements ✅ <br /> (primary) | - | - | 0 |
| Improvements ✅ <br /> (secondary) | - | - | 0 |
| All ❌✅ (primary) | 3.1% | [3.0%, 3.3%] | 4 |
- Regressions seem legit but nothing in self profiling stands out as a likely culprit.
- Asked the author for more investigation.


Subtree update of `rust-analyzer` [#121581](https://github.com/rust-lang/rust/pull/121581) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=710048f790ebc9ebfac6039c6733570f8084c9de&end=43fdd4916d19f4004e23d422b5547637ad67ab21&stat=instructions:u)

| (instructions:u) | mean | range | count |
|:----------------------------------:|:----:|:------------:|:-----:|
| Regressions ❌ <br /> (primary) | 0.7% | [0.7%, 0.8%] | 4 |
| Regressions ❌ <br /> (secondary) | 0.4% | [0.2%, 0.7%] | 23 |
| Improvements ✅ <br /> (primary) | - | - | 0 |
| Improvements ✅ <br /> (secondary) | - | - | 0 |
| All ❌✅ (primary) | 0.7% | [0.7%, 0.8%] | 4 |
- This is all noise - regressions revert in the next run.


#### Improvements

Improve codegen diagnostic handling [#121129](https://github.com/rust-lang/rust/pull/121129) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=026b3b8e955e0571db39aa96fc9d7aba25cc4d66&end=f70f19fef41cfdda75c92f163434c29ad046cf09&stat=instructions:u)

| (instructions:u) | mean | range | count |
|:----------------------------------:|:-----:|:--------------:|:-----:|
| Regressions ❌ <br /> (primary) | - | - | 0 |
| Regressions ❌ <br /> (secondary) | - | - | 0 |
| Improvements ✅ <br /> (primary) | - | - | 0 |
| Improvements ✅ <br /> (secondary) | -5.1% | [-5.1%, -5.1%] | 1 |
| All ❌✅ (primary) | - | - | 0 |


remove `sub_relations` from the `InferCtxt` [#119989](https://github.com/rust-lang/rust/pull/119989) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=d9ae43c7879c83f08ea2e886b55ea0f6839c9ead&end=397937d812852f9bbeb671005cb399dbcb357cde&stat=instructions:u)

| (instructions:u) | mean | range | count |
|:----------------------------------:|:-----:|:--------------:|:-----:|
| Regressions ❌ <br /> (primary) | - | - | 0 |
| Regressions ❌ <br /> (secondary) | - | - | 0 |
| Improvements ✅ <br /> (primary) | -0.3% | [-0.5%, -0.2%] | 28 |
| Improvements ✅ <br /> (secondary) | -0.4% | [-0.7%, -0.2%] | 28 |
| All ❌✅ (primary) | -0.3% | [-0.5%, -0.2%] | 28 |


Rollup of 7 pull requests [#121549](https://github.com/rust-lang/rust/pull/121549) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=6bdb8a4a96945527677f4b3dc7161875834189a4&end=381d69953bb7c3390cec0fee200f24529cb6320f&stat=instructions:u)

| (instructions:u) | mean | range | count |
|:----------------------------------:|:-----:|:--------------:|:-----:|
| Regressions ❌ <br /> (primary) | - | - | 0 |
| Regressions ❌ <br /> (secondary) | - | - | 0 |
| Improvements ✅ <br /> (primary) | -2.8% | [-2.8%, -2.8%] | 1 |
| Improvements ✅ <br /> (secondary) | - | - | 0 |
| All ❌✅ (primary) | -2.8% | [-2.8%, -2.8%] | 1 |


Use `br` instead of a conditional when switching on a constant boolean [#120650](https://github.com/rust-lang/rust/pull/120650) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=2ae1bb671183a072b54ed8ed39abfcd72990a3e7&end=89d8e3116ccb7f55760825b731de4f4e0fd76770&stat=instructions:u)

| (instructions:u) | mean | range | count |
|:----------------------------------:|:-----:|:--------------:|:-----:|
| Regressions ❌ <br /> (primary) | - | - | 0 |
| Regressions ❌ <br /> (secondary) | - | - | 0 |
| Improvements ✅ <br /> (primary) | -0.9% | [-2.2%, -0.1%] | 19 |
| Improvements ✅ <br /> (secondary) | - | - | 0 |
| All ❌✅ (primary) | -0.9% | [-2.2%, -0.1%] | 19 |


speed up `x install` by skipping archiving and compression [#118724](https://github.com/rust-lang/rust/pull/118724) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=43fdd4916d19f4004e23d422b5547637ad67ab21&end=26cd5d862e22c013ecb3396b177d3af80e95c836&stat=instructions:u)

| (instructions:u) | mean | range | count |
|:----------------------------------:|:-----:|:--------------:|:-----:|
| Regressions ❌ <br /> (primary) | - | - | 0 |
| Regressions ❌ <br /> (secondary) | - | - | 0 |
| Improvements ✅ <br /> (primary) | -0.7% | [-0.8%, -0.6%] | 4 |
| Improvements ✅ <br /> (secondary) | -0.4% | [-0.7%, -0.2%] | 24 |
| All ❌✅ (primary) | -0.7% | [-0.8%, -0.6%] | 4 |


Use generic `NonZero` in tests. [#121461](https://github.com/rust-lang/rust/pull/121461) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=0ecbd0605770f45c9151715e66ba2b3cae367fcb&end=0250ef2571185b05701ed9d74fc904c17508a397&stat=instructions:u)

| (instructions:u) | mean | range | count |
|:----------------------------------:|:-----:|:--------------:|:-----:|
| Regressions ❌ <br /> (primary) | - | - | 0 |
| Regressions ❌ <br /> (secondary) | - | - | 0 |
| Improvements ✅ <br /> (primary) | -3.6% | [-3.6%, -3.6%] | 1 |
| Improvements ✅ <br /> (secondary) | - | - | 0 |
| All ❌✅ (primary) | -3.6% | [-3.6%, -3.6%] | 1 |


#### Mixed

Rollup of 8 pull requests [#121345](https://github.com/rust-lang/rust/pull/121345) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=2b43e75c98cc5ae32328c8b49657bcd882eb5e75&end=bb594538fc6e84213a6b8d5e165442570aa48923&stat=instructions:u)

| (instructions:u) | mean | range | count |
|:----------------------------------:|:-----:|:--------------:|:-----:|
| Regressions ❌ <br /> (primary) | 0.6% | [0.4%, 1.1%] | 9 |
| Regressions ❌ <br /> (secondary) | 1.7% | [1.7%, 1.7%] | 1 |
| Improvements ✅ <br /> (primary) | -0.4% | [-0.8%, -0.2%] | 6 |
| Improvements ✅ <br /> (secondary) | -0.8% | [-2.1%, -0.3%] | 17 |
| All ❌✅ (primary) | 0.2% | [-0.8%, 1.1%] | 15 |
- Lots of seemingly risky PRs in this roll up. Need to run through them.


match lowering: eagerly simplify match pairs [#120904](https://github.com/rust-lang/rust/pull/120904) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=2dc0170233689435f988579a2080f27b02027d21&end=4e650749337bd4b0b464ebb5252d122b7cb98c8d&stat=instructions:u)

| (instructions:u) | mean | range | count |
|:----------------------------------:|:-----:|:--------------:|:-----:|
| Regressions ❌ <br /> (primary) | - | - | 0 |
| Regressions ❌ <br /> (secondary) | 0.2% | [0.2%, 0.2%] | 3 |
| Improvements ✅ <br /> (primary) | -2.0% | [-3.1%, -0.2%] | 7 |
| Improvements ✅ <br /> (secondary) | - | - | 0 |
| All ❌✅ (primary) | -2.0% | [-3.1%, -0.2%] | 7 |
- Improvements far outweigh the regressions so I don't think this needs further investigation.


compiler: clippy::complexity fixes [#121523](https://github.com/rust-lang/rust/pull/121523) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=8f359beca4e58bc3ae795a666301a8f47023044c&end=6bdb8a4a96945527677f4b3dc7161875834189a4&stat=instructions:u)

| (instructions:u) | mean | range | count |
|:----------------------------------:|:-----:|:--------------:|:-----:|
| Regressions ❌ <br /> (primary) | 1.5% | [0.2%, 2.7%] | 2 |
| Regressions ❌ <br /> (secondary) | - | - | 0 |
| Improvements ✅ <br /> (primary) | - | - | 0 |
| Improvements ✅ <br /> (secondary) | -1.2% | [-1.2%, -1.2%] | 1 |
| All ❌✅ (primary) | 1.5% | [0.2%, 2.7%] | 2 |
- The large regression that caused this to be marked overall as a regression is just noise.


Add `#[rustc_no_mir_inline]` for standard library UB checks [#121114](https://github.com/rust-lang/rust/pull/121114) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=89d8e3116ccb7f55760825b731de4f4e0fd76770&end=e9f95949138125bb1b98ed213a41b9aa825bacf5&stat=instructions:u)

| (instructions:u) | mean | range | count |
|:----------------------------------:|:-----:|:--------------:|:-----:|
| Regressions ❌ <br /> (primary) | 0.5% | [0.2%, 1.2%] | 33 |
| Regressions ❌ <br /> (secondary) | 1.5% | [0.3%, 5.5%] | 12 |
| Improvements ✅ <br /> (primary) | -0.3% | [-0.3%, -0.3%] | 3 |
| Improvements ✅ <br /> (secondary) | -0.3% | [-0.3%, -0.2%] | 8 |
| All ❌✅ (primary) | 0.4% | [-0.3%, 1.2%] | 36 |
- Consider necessary for fixing some pathological performance cases as well as addressing another critical issue. (See [this analysis](https://github.com/rust-lang/rust/pull/121114#issuecomment-1962316858) for more detail.)


Implement RFC 3373: Avoid non-local definitions in functions [#120393](https://github.com/rust-lang/rust/pull/120393) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=8c0b1fcd2914caaf1c3a1071028fb74b70c519e9&end=b0d3e04ca9f848c5fce3ad4bab6153009be06db0&stat=instructions:u)

| (instructions:u) | mean | range | count |
|:----------------------------------:|:-----:|:--------------:|:-----:|
| Regressions ❌ <br /> (primary) | 2.2% | [0.5%, 4.2%] | 12 |
| Regressions ❌ <br /> (secondary) | 0.7% | [0.1%, 1.8%] | 5 |
| Improvements ✅ <br /> (primary) | - | - | 0 |
| Improvements ✅ <br /> (secondary) | -0.4% | [-0.8%, -0.1%] | 27 |
| All ❌✅ (primary) | 2.2% | [0.5%, 4.2%] | 12 |
- [#121625](https://github.com/rust-lang/rust/pull/121625) is meant to try to address this.
- Given the complex nature of the change, I imagine a revert is not desirable even if performance can't be gained back.