Skip to content

Commit 7a1e00a

Browse files
KobzolKobzol
authored
Merge pull request #2153 from panstromek/patch-5
add 2025-06-09 perf triage
2 parents 6cc9025 + 85e22cd commit 7a1e00a

File tree

1 file changed

+197
-0
lines changed

1 file changed

+197
-0
lines changed

‎triage/2025-06-09.md

Lines changed: 197 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,197 @@
1+
# 2025-06-09 Triage Log
2+
3+
Mostly positive week, with a lot of improvements in the type system, especially in new solver and one big win in caching code. Regressions come from new warnings, with outsized impact on one benchmark with a lot of generated code.
4+
5+
Triage done by **@panstromek**.
6+
Revision range: [2fc3deed..c31cccb7](https://perf.rust-lang.org/?start=2fc3deed9fcb8762ad57191e0195f06f7543e4a5&end=c31cccb7b5cc098b1a8c1794ed38d7fdbec0ccb0&absolute=false&stat=instructions%3Au)
7+
8+
**Summary**:
9+
10+
| (instructions:u) | mean | range | count |
11+
|:----------------------------------:|:-----:|:---------------:|:-----:|
12+
| Regressions ❌ <br /> (primary) | 3.1% | [0.3%, 8.5%] | 22 |
13+
| Regressions ❌ <br /> (secondary) | 0.6% | [0.2%, 0.9%] | 3 |
14+
| Improvements ✅ <br /> (primary) | -1.0% | [-3.4%, -0.2%] | 151 |
15+
| Improvements ✅ <br /> (secondary) | -3.5% | [-66.5%, -0.2%] | 146 |
16+
| All ❌✅ (primary) | -0.4% | [-3.4%, 8.5%] | 173 |
17+
18+
19+
3 Regressions, 6 Improvements, 5 Mixed; 2 of them in rollups
20+
45 artifact comparisons made in total
21+
22+
#### Regressions
23+
24+
Use ccache for stage0 tool builds [#136942](https://github.com/rust-lang/rust/pull/136942) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=2f176126aaf3fd316fc5ced3d9a111a48797b52c&end=aae43c4532690153af7465227816c93036bb1604&stat=instructions:u)
25+
26+
| (instructions:u) | mean | range | count |
27+
|:----------------------------------:|:----:|:------------:|:-----:|
28+
| Regressions ❌ <br /> (primary) | 2.9% | [2.9%, 2.9%] | 1 |
29+
| Regressions ❌ <br /> (secondary) | - | - | 0 |
30+
| Improvements ✅ <br /> (primary) | - | - | 0 |
31+
| Improvements ✅ <br /> (secondary) | - | - | 0 |
32+
| All ❌✅ (primary) | 2.9% | [2.9%, 2.9%] | 1 |
33+
34+
The regression is spurious, the affected benchmark is bi-modal.
35+
36+
Move placeholder handling to a proper preprocessing step [#140466](https://github.com/rust-lang/rust/pull/140466) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=81a964c23ea4fe9ab52b4449bb166bf280035797&end=425e142686242c7e73f5e32c79071ae266f0f355&stat=instructions:u)
37+
38+
| (instructions:u) | mean | range | count |
39+
|:----------------------------------:|:----:|:------------:|:-----:|
40+
| Regressions ❌ <br /> (primary) | 0.1% | [0.1%, 0.1%] | 1 |
41+
| Regressions ❌ <br /> (secondary) | 0.2% | [0.2%, 0.3%] | 8 |
42+
| Improvements ✅ <br /> (primary) | - | - | 0 |
43+
| Improvements ✅ <br /> (secondary) | - | - | 0 |
44+
| All ❌✅ (primary) | 0.1% | [0.1%, 0.1%] | 1 |
45+
46+
47+
These regressions look real. One tiny regression in Cargo and two secondary benchmarks - `coercions` and `ucd`, which are both based on giant static tables and are somewhat artificial.
48+
49+
According to a [comment by author](https://github.com/rust-lang/rust/pull/140466#issuecomment-2955437404), this PR doesn't change runtime behavour, so those are probably just spurious. Most of the benchmarks improved since then and some of them are a bit noisy at times.
50+
51+
This is a piece of larger important work (Polonius) and the regressions are pretty limited. [#130227](https://github.com/rust-lang/rust/pull/130227) (which this PR is part of) does have a bit different perf results, so I guess it will also make more sense asses the perf impact of this work when more pieces of it lands.
52+
53+
Add a new `mismatched-lifetime-syntaxes` lint [#138677](https://github.com/rust-lang/rust/pull/138677) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=076ec59ff1dcf538b9d3a0b8e0d7f4edd0559959&end=ccf3198de316b488ee17441935182e9d5292b4d3&stat=instructions:u)
54+
55+
| (instructions:u) | mean | range | count |
56+
|:----------------------------------:|:-----:|:--------------:|:-----:|
57+
| Regressions ❌ <br /> (primary) | 1.6% | [0.2%, 9.7%] | 72 |
58+
| Regressions ❌ <br /> (secondary) | 0.3% | [0.2%, 0.5%] | 11 |
59+
| Improvements ✅ <br /> (primary) | - | - | 0 |
60+
| Improvements ✅ <br /> (secondary) | -0.1% | [-0.1%, -0.1%] | 1 |
61+
| All ❌✅ (primary) | 1.6% | [0.2%, 9.7%] | 72 |
62+
63+
Large regression on many benchmarks, most notably on `stm32f4`. Mostly caused by emitting more warnings. Author justified the regression in [an comment on the PR](https://github.com/rust-lang/rust/pull/138677#issuecomment-2940461051). Results and next steps are discussed in [dedicated zulip channel](https://rust-lang.zulipchat.com/#narrow/stream/131828-t-compiler/topic/Perf.20regression.20for.20warn-by-default.20lint).
64+
65+
#### Improvements
66+
67+
Change per-module naked fn checks to happen during typeck instead [#141774](https://github.com/rust-lang/rust/pull/141774) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=9f0e5d963d05ebcf7ae1ca1f4bda3668c702acda&end=44f415c1d617ebc7b931a243b7b321ef8a6ca47c&stat=instructions:u)
68+
69+
| (instructions:u) | mean | range | count |
70+
|:----------------------------------:|:-----:|:--------------:|:-----:|
71+
| Regressions ❌ <br /> (primary) | - | - | 0 |
72+
| Regressions ❌ <br /> (secondary) | - | - | 0 |
73+
| Improvements ✅ <br /> (primary) | -0.1% | [-0.1%, -0.1%] | 11 |
74+
| Improvements ✅ <br /> (secondary) | -0.3% | [-0.4%, -0.2%] | 11 |
75+
| All ❌✅ (primary) | -0.1% | [-0.1%, -0.1%] | 11 |
76+
77+
78+
Rollup of 11 pull requests [#142133](https://github.com/rust-lang/rust/pull/142133) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=868bf2da31d1c6ead7016d5ee52eacfd2e2fd158&end=775e0c8aeb8f63192854b27156f8b05a06b51814&stat=instructions:u)
79+
80+
| (instructions:u) | mean | range | count |
81+
|:----------------------------------:|:-----:|:--------------:|:-----:|
82+
| Regressions ❌ <br /> (primary) | - | - | 0 |
83+
| Regressions ❌ <br /> (secondary) | - | - | 0 |
84+
| Improvements ✅ <br /> (primary) | -0.2% | [-0.2%, -0.2%] | 4 |
85+
| Improvements ✅ <br /> (secondary) | -0.3% | [-0.4%, -0.2%] | 8 |
86+
| All ❌✅ (primary) | -0.2% | [-0.2%, -0.2%] | 4 |
87+
88+
89+
Clear nested candidates in select if certainty is yes [#141927](https://github.com/rust-lang/rust/pull/141927) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=5e0bdaa9dde845b8e44fd93bf0c09d21ca60daa1&end=2f2c8c3512e82e4315db83bbb53eb79e2c566270&stat=instructions:u)
90+
91+
| (instructions:u) | mean | range | count |
92+
|:----------------------------------:|:-----:|:---------------:|:-----:|
93+
| Regressions ❌ <br /> (primary) | - | - | 0 |
94+
| Regressions ❌ <br /> (secondary) | - | - | 0 |
95+
| Improvements ✅ <br /> (primary) | - | - | 0 |
96+
| Improvements ✅ <br /> (secondary) | -3.5% | [-20.7%, -0.4%] | 11 |
97+
| All ❌✅ (primary) | - | - | 0 |
98+
99+
100+
Remove CollectItemTypesVisitor [#142074](https://github.com/rust-lang/rust/pull/142074) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=a5584a8fe16037dc01782064fa41424a6dbe9987&end=0b65d0db5f77c44bf37b4a25489562d68c14aecf&stat=instructions:u)
101+
102+
| (instructions:u) | mean | range | count |
103+
|:----------------------------------:|:-----:|:--------------:|:-----:|
104+
| Regressions ❌ <br /> (primary) | - | - | 0 |
105+
| Regressions ❌ <br /> (secondary) | - | - | 0 |
106+
| Improvements ✅ <br /> (primary) | -0.1% | [-0.1%, -0.1%] | 1 |
107+
| Improvements ✅ <br /> (secondary) | -0.2% | [-0.4%, -0.1%] | 8 |
108+
| All ❌✅ (primary) | -0.1% | [-0.1%, -0.1%] | 1 |
109+
110+
111+
Filter out universals and lifetimes from `stalled_vars` [#142088](https://github.com/rust-lang/rust/pull/142088) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=8072811356a178dbdf8ca09b1635cfafd4661971&end=244bbfc60ee8593db96892468eee876240cb7ba1&stat=instructions:u)
112+
113+
| (instructions:u) | mean | range | count |
114+
|:----------------------------------:|:------:|:---------------:|:-----:|
115+
| Regressions ❌ <br /> (primary) | - | - | 0 |
116+
| Regressions ❌ <br /> (secondary) | - | - | 0 |
117+
| Improvements ✅ <br /> (primary) | - | - | 0 |
118+
| Improvements ✅ <br /> (secondary) | -10.2% | [-29.8%, -0.1%] | 15 |
119+
| All ❌✅ (primary) | - | - | 0 |
120+
121+
122+
Simplify and optimize `VecCache`'s `SlotIndex::from_index` [#142095](https://github.com/rust-lang/rust/pull/142095) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=244bbfc60ee8593db96892468eee876240cb7ba1&end=fb644e6a1a7d34c6bbb5ecfe5c185f8c977d6bb3&stat=instructions:u)
123+
124+
| (instructions:u) | mean | range | count |
125+
|:----------------------------------:|:-----:|:--------------:|:-----:|
126+
| Regressions ❌ <br /> (primary) | - | - | 0 |
127+
| Regressions ❌ <br /> (secondary) | 0.3% | [0.2%, 0.4%] | 4 |
128+
| Improvements ✅ <br /> (primary) | -0.8% | [-3.4%, -0.2%] | 184 |
129+
| Improvements ✅ <br /> (secondary) | -1.0% | [-4.8%, -0.2%] | 112 |
130+
| All ❌✅ (primary) | -0.8% | [-3.4%, -0.2%] | 184 |
131+
132+
133+
#### Mixed
134+
135+
Merge coroutine obligation checking into borrowck parallel loop [#141773](https://github.com/rust-lang/rust/pull/141773) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=2fc3deed9fcb8762ad57191e0195f06f7543e4a5&end=91fad92585b2dafc52a074e502b2a6c1f093ca35&stat=instructions:u)
136+
137+
| (instructions:u) | mean | range | count |
138+
|:----------------------------------:|:-----:|:--------------:|:-----:|
139+
| Regressions ❌ <br /> (primary) | - | - | 0 |
140+
| Regressions ❌ <br /> (secondary) | 0.4% | [0.3%, 0.5%] | 2 |
141+
| Improvements ✅ <br /> (primary) | -2.9% | [-2.9%, -2.9%] | 1 |
142+
| Improvements ✅ <br /> (secondary) | - | - | 0 |
143+
| All ❌✅ (primary) | -2.9% | [-2.9%, -2.9%] | 1 |
144+
145+
Looks like noise, the affected benchmarks keep following similar pattern (`clap-derive` is bimodal, the other two secondary benchmarks are noisy).
146+
147+
Rollup of 8 pull requests [#141944](https://github.com/rust-lang/rust/pull/141944) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=b17dba45186c454576d0fc8fb93ecc65eb1a763a&end=c68032fd4c442d275f4daa571ba19c076106b490&stat=instructions:u)
148+
149+
| (instructions:u) | mean | range | count |
150+
|:----------------------------------:|:-----:|:--------------:|:-----:|
151+
| Regressions ❌ <br /> (primary) | - | - | 0 |
152+
| Regressions ❌ <br /> (secondary) | 0.2% | [0.2%, 0.3%] | 5 |
153+
| Improvements ✅ <br /> (primary) | -0.6% | [-1.1%, -0.2%] | 3 |
154+
| Improvements ✅ <br /> (secondary) | - | - | 0 |
155+
| All ❌✅ (primary) | -0.6% | [-1.1%, -0.2%] | 3 |
156+
157+
Few regressions in unused-warnings. This looks like noise to me for the most part - check and opt builds don't agree on whether frontend times regressed or not and the benchmark since returned to previous state (partly in another rollup).
158+
159+
I would suspect #141741 to be the cause, because it touched use statements and unused-warnings is a large list of use statements, but that one had different pre-merge results. I don't think this is worth digging into, though.
160+
161+
Rework `collect_and_apply` to not rely on size hint for optimization [#141652](https://github.com/rust-lang/rust/pull/141652) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=792fc2b033aea7ea7b766e38bdc40f7d6bdce8c3&end=d9a739363878a06633eb912e44f02a3c5d6ba08c&stat=instructions:u)
162+
163+
| (instructions:u) | mean | range | count |
164+
|:----------------------------------:|:-----:|:--------------:|:-----:|
165+
| Regressions ❌ <br /> (primary) | - | - | 0 |
166+
| Regressions ❌ <br /> (secondary) | 0.3% | [0.3%, 0.4%] | 6 |
167+
| Improvements ✅ <br /> (primary) | -0.2% | [-0.4%, -0.1%] | 9 |
168+
| Improvements ✅ <br /> (secondary) | -0.5% | [-1.0%, -0.2%] | 49 |
169+
| All ❌✅ (primary) | -0.2% | [-0.4%, -0.1%] | 9 |
170+
171+
Improvements outweigh regressions. Regressions are limited to `wg-grammar`. Also justified by [review comment](https://github.com/rust-lang/rust/pull/141652#issuecomment-2936239135).
172+
173+
Fast path for stalled obligations on self ty [#141681](https://github.com/rust-lang/rust/pull/141681) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=f315e6145802e091ff9fceab6db627a4b4ec2b86&end=9f0e5d963d05ebcf7ae1ca1f4bda3668c702acda&stat=instructions:u)
174+
175+
| (instructions:u) | mean | range | count |
176+
|:----------------------------------:|:-----:|:---------------:|:-----:|
177+
| Regressions ❌ <br /> (primary) | 0.2% | [0.2%, 0.2%] | 1 |
178+
| Regressions ❌ <br /> (secondary) | 0.6% | [0.1%, 1.3%] | 14 |
179+
| Improvements ✅ <br /> (primary) | -0.2% | [-0.3%, -0.1%] | 25 |
180+
| Improvements ✅ <br /> (secondary) | -4.3% | [-50.3%, -0.0%] | 43 |
181+
| All ❌✅ (primary) | -0.2% | [-0.3%, 0.2%] | 26 |
182+
183+
Improvements outweigh regressions, those are mostly on stress tests. Some regressions also don't seem to be related, they just return back to state before [#141652](https://github.com/rust-lang/rust/pull/141652).
184+
185+
Don't walk into `Certainty::Yes` goals [#142085](https://github.com/rust-lang/rust/pull/142085) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=0b65d0db5f77c44bf37b4a25489562d68c14aecf&end=8072811356a178dbdf8ca09b1635cfafd4661971&stat=instructions:u)
186+
187+
| (instructions:u) | mean | range | count |
188+
|:----------------------------------:|:-----:|:---------------:|:-----:|
189+
| Regressions ❌ <br /> (primary) | 1.4% | [1.4%, 1.4%] | 1 |
190+
| Regressions ❌ <br /> (secondary) | 1.4% | [1.4%, 1.4%] | 1 |
191+
| Improvements ✅ <br /> (primary) | - | - | 0 |
192+
| Improvements ✅ <br /> (secondary) | -8.7% | [-17.9%, -0.8%] | 6 |
193+
| All ❌✅ (primary) | 1.4% | [1.4%, 1.4%] | 1 |
194+
195+
Syn regression is spurious, it's just returning to the default state in bimodal pattern.
196+
197+
Secondary regression is limited to new solver benchmark. New solver is not enabled by default, yet.

0 commit comments

Comments
 (0)