Skip to content

Commit 804d841

Browse files
lqdlqd
authored
Merge pull request #2481 from rust-lang/panstromek-patch-1
Add triage for 2026-05-25
2 parents fd5f144 + 4e9e90e commit 804d841

1 file changed

Lines changed: 135 additions & 0 deletions

File tree

‎triage/2026/2026-05-25.md‎

Lines changed: 135 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,135 @@
1+
# 2026-05-25 Triage Log
2+
3+
This week was largely positive, with most of the improvements coming from algorithm change in visibility checking: [#156228](https://github.com/rust-lang/rust/pull/156228).
4+
5+
Triage done by **@panstromek**.
6+
Revision range: [281c97c3..783eb8c8](https://perf.rust-lang.org/?start=281c97c3240a9abd984ca0c6a2cd7389115e80d5&end=783eb8c8682ddde0807c60ed8293670ef523794f&absolute=false&stat=instructions%3Au)
7+
8+
**Summary**:
9+
10+
| (instructions:u) | mean | range | count |
11+
|:----------------------------------:|:-----:|:--------------:|:-----:|
12+
| Regressions ❌ <br /> (primary) | 0.4% | [0.1%, 0.7%] | 5 |
13+
| Regressions ❌ <br /> (secondary) | 0.5% | [0.1%, 1.1%] | 16 |
14+
| Improvements ✅ <br /> (primary) | -0.9% | [-6.6%, -0.1%] | 164 |
15+
| Improvements ✅ <br /> (secondary) | -0.4% | [-1.3%, -0.1%] | 51 |
16+
| All ❌✅ (primary) | -0.9% | [-6.6%, 0.7%] | 169 |
17+
18+
19+
2 Regressions, 2 Improvements, 5 Mixed; 2 of them in rollups
20+
34 artifact comparisons made in total
21+
22+
#### Regressions
23+
24+
Stabilize `--remap-path-prefix` in rustdoc [#155307](https://github.com/rust-lang/rust/pull/155307) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=e96c36b6f76833388c519561d145492d2c08db4e&end=4d276d7fdba2a24c73dbca3027461bb202e83436&stat=instructions:u)
25+
26+
| (instructions:u) | mean | range | count |
27+
|:----------------------------------:|:----:|:------------:|:-----:|
28+
| Regressions ❌ <br /> (primary) | 0.7% | [0.7%, 0.7%] | 1 |
29+
| Regressions ❌ <br /> (secondary) | 0.4% | [0.4%, 0.4%] | 1 |
30+
| Improvements ✅ <br /> (primary) | - | - | 0 |
31+
| Improvements ✅ <br /> (secondary) | - | - | 0 |
32+
| All ❌✅ (primary) | 0.7% | [0.7%, 0.7%] | 1 |
33+
34+
`nalgebra` regressions are noise (return to default state after previous PR).
35+
36+
Rollup of 4 pull requests [#156881](https://github.com/rust-lang/rust/pull/156881) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=1d59f669f9c699b982f112ebbc81e8b1eb550d68&end=609b8c5cefb3932bbaf4497cb7f9195ca8a1eab6&stat=instructions:u)
37+
38+
| (instructions:u) | mean | range | count |
39+
|:----------------------------------:|:-----:|:--------------:|:-----:|
40+
| Regressions ❌ <br /> (primary) | 0.3% | [0.2%, 0.4%] | 11 |
41+
| Regressions ❌ <br /> (secondary) | 0.2% | [0.0%, 0.6%] | 12 |
42+
| Improvements ✅ <br /> (primary) | - | - | 0 |
43+
| Improvements ✅ <br /> (secondary) | -1.9% | [-1.9%, -1.9%] | 1 |
44+
| All ❌✅ (primary) | 0.3% | [0.2%, 0.4%] | 11 |
45+
46+
Looks to be caused by https://github.com/rust-lang/rust/pull/156824. I couldn't confirm this directly because we miss the unrolled build, but looking at the other PRs, one is a pure test change, and perf runs on the other two came back with no relevant changes, so this is the only one remaning.
47+
48+
It doesn't seem like that big of a deal to me (mostly tiny benchmarks, `incr-unchanged/patched`), but seems unexpected. Some of them regress in queries that don't seem related (match checking). Left a comment there.
49+
50+
#### Improvements
51+
52+
Privacy: try use queue instead of fixed-point iteration [#156228](https://github.com/rust-lang/rust/pull/156228) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=ca9203f29c609b344566d4eed9b5d9a3bf10290c&end=60592ad7c63cdb7cfe092f785bb224b655512926&stat=instructions:u)
53+
54+
| (instructions:u) | mean | range | count |
55+
|:----------------------------------:|:-----:|:--------------:|:-----:|
56+
| Regressions ❌ <br /> (primary) | - | - | 0 |
57+
| Regressions ❌ <br /> (secondary) | - | - | 0 |
58+
| Improvements ✅ <br /> (primary) | -0.9% | [-6.7%, -0.1%] | 144 |
59+
| Improvements ✅ <br /> (secondary) | -0.5% | [-1.3%, -0.1%] | 31 |
60+
| All ❌✅ (primary) | -0.9% | [-6.7%, -0.1%] | 144 |
61+
62+
63+
Update LLVM to 22.1.6 [#156757](https://github.com/rust-lang/rust/pull/156757) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=037b621e831cba9ebd4148f9de33eff361df23d3&end=e96c36b6f76833388c519561d145492d2c08db4e&stat=instructions:u)
64+
65+
| (instructions:u) | mean | range | count |
66+
|:----------------------------------:|:-----:|:--------------:|:-----:|
67+
| Regressions ❌ <br /> (primary) | - | - | 0 |
68+
| Regressions ❌ <br /> (secondary) | - | - | 0 |
69+
| Improvements ✅ <br /> (primary) | -0.7% | [-0.7%, -0.7%] | 1 |
70+
| Improvements ✅ <br /> (secondary) | -0.4% | [-0.4%, -0.4%] | 1 |
71+
| All ❌✅ (primary) | -0.7% | [-0.7%, -0.7%] | 1 |
72+
73+
note: this result is spurious, we got inverse results in the next PR ([#155307](https://github.com/rust-lang/rust/pull/155307), first PR in the regressions list).
74+
75+
#### Mixed
76+
77+
Make bitset `would_modify_words` more vectorizer-friendly [#153640](https://github.com/rust-lang/rust/pull/153640) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=1ea1171c1e537f295225be1c7b67dba46794e6ad&end=e50aa6fba4e63ab34c72bf9acfd2c307c1155d1a&stat=instructions:u)
78+
79+
| (instructions:u) | mean | range | count |
80+
|:----------------------------------:|:-----:|:--------------:|:-----:|
81+
| Regressions ❌ <br /> (primary) | - | - | 0 |
82+
| Regressions ❌ <br /> (secondary) | 0.1% | [0.1%, 0.1%] | 1 |
83+
| Improvements ✅ <br /> (primary) | -1.4% | [-2.2%, -0.7%] | 4 |
84+
| Improvements ✅ <br /> (secondary) | - | - | 0 |
85+
| All ❌✅ (primary) | -1.4% | [-2.2%, -0.7%] | 4 |
86+
87+
`include-blob` regression is noise.
88+
89+
Fix jump-to-def links broken by turbofish syntax [#156727](https://github.com/rust-lang/rust/pull/156727) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=4b9792692fbb675174d4d2082e7c37b2bc930e71&end=1f8e04d34ab0c1fd9574840aa6db670e41593bfb&stat=instructions:u)
90+
91+
| (instructions:u) | mean | range | count |
92+
|:----------------------------------:|:-----:|:--------------:|:-----:|
93+
| Regressions ❌ <br /> (primary) | 0.7% | [0.7%, 0.7%] | 1 |
94+
| Regressions ❌ <br /> (secondary) | 0.4% | [0.4%, 0.4%] | 1 |
95+
| Improvements ✅ <br /> (primary) | - | - | 0 |
96+
| Improvements ✅ <br /> (secondary) | -0.1% | [-0.1%, -0.1%] | 1 |
97+
| All ❌✅ (primary) | 0.7% | [0.7%, 0.7%] | 1 |
98+
99+
Perf results are noise, this PR doesn't change anything in the compiler. It seems that `nalgebra` `opt` became bimodal.
100+
101+
Rollup of 2 pull requests [#156794](https://github.com/rust-lang/rust/pull/156794) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=b954122bbe6c4a244c511a9a168c6d5438d8f58f&end=037b621e831cba9ebd4148f9de33eff361df23d3&stat=instructions:u)
102+
103+
| (instructions:u) | mean | range | count |
104+
|:----------------------------------:|:-----:|:--------------:|:-----:|
105+
| Regressions ❌ <br /> (primary) | 0.3% | [0.2%, 0.4%] | 64 |
106+
| Regressions ❌ <br /> (secondary) | 0.3% | [0.1%, 0.4%] | 38 |
107+
| Improvements ✅ <br /> (primary) | -0.3% | [-0.5%, -0.2%] | 9 |
108+
| Improvements ✅ <br /> (secondary) | -0.2% | [-0.2%, -0.2%] | 8 |
109+
| All ❌✅ (primary) | 0.2% | [-0.5%, 0.4%] | 73 |
110+
111+
Caused by https://github.com/rust-lang/rust/pull/156242, already triaged by @JonathanBrouwer, addressed by @Saethlin in https://github.com/rust-lang/rust/pull/156849.
112+
113+
Implement fast path for `derive(PartialOrd)` when deriving `Ord` [#155598](https://github.com/rust-lang/rust/pull/155598) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=23a3312d92a1c4ba0373f1e25277be20ba8bb28c&end=dd8b2d6162c5616d7030238c29f892a431093cdd&stat=instructions:u)
114+
115+
| (instructions:u) | mean | range | count |
116+
|:----------------------------------:|:-----:|:--------------:|:-----:|
117+
| Regressions ❌ <br /> (primary) | 0.7% | [0.5%, 0.8%] | 3 |
118+
| Regressions ❌ <br /> (secondary) | 0.7% | [0.2%, 1.0%] | 9 |
119+
| Improvements ✅ <br /> (primary) | -0.3% | [-0.5%, -0.2%] | 20 |
120+
| Improvements ✅ <br /> (secondary) | -0.3% | [-0.5%, -0.2%] | 5 |
121+
| All ❌✅ (primary) | -0.2% | [-0.5%, 0.8%] | 23 |
122+
123+
Improvements mostly outweigh regressions. Results roughly match pre-merge results, so I assume the comments above still hold and this is considered fine.
124+
125+
Re-add call site inlining attributes [#156849](https://github.com/rust-lang/rust/pull/156849) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=a06c1ca53a775af31bcbdaeed995c2f224b34d06&end=1d59f669f9c699b982f112ebbc81e8b1eb550d68&stat=instructions:u)
126+
127+
| (instructions:u) | mean | range | count |
128+
|:----------------------------------:|:-----:|:--------------:|:-----:|
129+
| Regressions ❌ <br /> (primary) | 0.3% | [0.3%, 0.4%] | 2 |
130+
| Regressions ❌ <br /> (secondary) | 0.2% | [0.2%, 0.2%] | 8 |
131+
| Improvements ✅ <br /> (primary) | -0.3% | [-0.4%, -0.2%] | 66 |
132+
| Improvements ✅ <br /> (secondary) | -0.3% | [-0.5%, -0.2%] | 42 |
133+
| All ❌✅ (primary) | -0.3% | [-0.4%, 0.4%] | 68 |
134+
135+
Fixes the regression from https://github.com/rust-lang/rust/pull/156242, already triaged by @saethlin (author): "Perf impact is the inverse of the PR linked in the description, within typical jitter. The net effect of the two PRs is nothing."

0 commit comments

Comments
 (0)