Skip to content

Commit 5972835

Browse files
Mark-SimulacrumMark-Simulacrum
committed
Add triage for this week
1 parent c1abcdb commit 5972835

File tree

1 file changed

+123
-0
lines changed

1 file changed

+123
-0
lines changed

‎triage/2024-04-29.md‎

Lines changed: 123 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,123 @@
1+
# 2024-04-29 Triage Log
2+
3+
Several non-noise changes this week, with both improvements and regresions
4+
coming as a result. Overall compiler performance is roughly neutral across the
5+
week.
6+
7+
Triage done by **@simulacrum**.
8+
Revision range: [a77f76e2..c65b2dc9](https://perf.rust-lang.org/?start=a77f76e26302e9a084fb321817675b1dfc1dcd63&end=c65b2dc935c27c0c8c3997c6e8d8894718a2cb1a&absolute=false&stat=instructions%3Au)
9+
10+
**Summary**:
11+
12+
| (instructions:u) | mean | range | count |
13+
|:----------------------------------:|:-----:|:---------------:|:-----:|
14+
| Regressions ❌ <br /> (primary) | 0.4% | [0.2%, 1.4%] | 104 |
15+
| Regressions ❌ <br /> (secondary) | 2.4% | [0.2%, 23.7%] | 81 |
16+
| Improvements ✅ <br /> (primary) | -3.8% | [-26.1%, -0.3%] | 10 |
17+
| Improvements ✅ <br /> (secondary) | -1.6% | [-4.6%, -0.5%] | 12 |
18+
| All ❌✅ (primary) | 0.1% | [-26.1%, 1.4%] | 114 |
19+
20+
21+
2 Regressions, 2 Improvements, 3 Mixed; 1 of them in rollups
22+
51 artifact comparisons made in total
23+
24+
#### Regressions
25+
26+
Use fulfillment in method probe, not evaluation [#122317](https://github.com/rust-lang/rust/pull/122317) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=ad07aa12c99698f810d730d7b5a49704c729651d&end=cd90d5c03532da6f7ca7dcfb861ffabdc36a9d00&stat=instructions:u)
27+
28+
| (instructions:u) | mean | range | count |
29+
|:----------------------------------:|:----:|:-------------:|:-----:|
30+
| Regressions ❌ <br /> (primary) | 0.5% | [0.2%, 1.3%] | 38 |
31+
| Regressions ❌ <br /> (secondary) | 4.2% | [0.5%, 23.6%] | 39 |
32+
| Improvements ✅ <br /> (primary) | - | - | 0 |
33+
| Improvements ✅ <br /> (secondary) | - | - | 0 |
34+
| All ❌✅ (primary) | 0.5% | [0.2%, 1.3%] | 38 |
35+
36+
Some additional attempts to fix perf were done in a follow-up PR
37+
(https://github.com/rust-lang/rust/pull/124303) but did not pan out to be
38+
successful. It looks like this is a correctness fix, so we'll need to accept
39+
the regressions for now.
40+
41+
Stabilize `Utf8Chunks` [#123909](https://github.com/rust-lang/rust/pull/123909) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=1b3a32958bb54bde45e693217e8f7469459e5865&end=4d570eea025a19564429eb52b34ec34e14659f55&stat=instructions:u)
42+
43+
| (instructions:u) | mean | range | count |
44+
|:----------------------------------:|:----:|:------------:|:-----:|
45+
| Regressions ❌ <br /> (primary) | 0.5% | [0.2%, 1.1%] | 11 |
46+
| Regressions ❌ <br /> (secondary) | 0.8% | [0.3%, 1.2%] | 19 |
47+
| Improvements ✅ <br /> (primary) | - | - | 0 |
48+
| Improvements ✅ <br /> (secondary) | - | - | 0 |
49+
| All ❌✅ (primary) | 0.5% | [0.2%, 1.1%] | 11 |
50+
51+
The regressions are in doc builds, but are not really expected from what is a
52+
relatively small change. Further investigation feels warranted (see
53+
https://github.com/rust-lang/rust/pull/123909#issuecomment-2082668500).
54+
55+
#### Improvements
56+
57+
Rollup of 5 pull requests [#124289](https://github.com/rust-lang/rust/pull/124289) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=c67277301c896857d0534f2bb7431680796833fb&end=ad07aa12c99698f810d730d7b5a49704c729651d&stat=instructions:u)
58+
59+
| (instructions:u) | mean | range | count |
60+
|:----------------------------------:|:-----:|:--------------:|:-----:|
61+
| Regressions ❌ <br /> (primary) | - | - | 0 |
62+
| Regressions ❌ <br /> (secondary) | - | - | 0 |
63+
| Improvements ✅ <br /> (primary) | -0.3% | [-0.3%, -0.2%] | 7 |
64+
| Improvements ✅ <br /> (secondary) | -0.2% | [-0.2%, -0.2%] | 2 |
65+
| All ❌✅ (primary) | -0.3% | [-0.3%, -0.2%] | 7 |
66+
67+
Unclear whether this is a genuine improvement. Performance seems to have
68+
re-regressed in #123126 (see Mixed results below), so this may just be
69+
bimodality of some kind.
70+
71+
Set writable and dead_on_unwind attributes for sret arguments [#121298](https://github.com/rust-lang/rust/pull/121298) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=cb3752d20e0f5d24348062211102a08d46fbecff&end=284f94f9c0f77ad4ef85323a634cfda29c1a801d&stat=instructions:u)
72+
73+
| (instructions:u) | mean | range | count |
74+
|:----------------------------------:|:-----:|:---------------:|:-----:|
75+
| Regressions ❌ <br /> (primary) | 0.5% | [0.5%, 0.5%] | 1 |
76+
| Regressions ❌ <br /> (secondary) | - | - | 0 |
77+
| Improvements ✅ <br /> (primary) | -3.1% | [-26.0%, -0.3%] | 12 |
78+
| Improvements ✅ <br /> (secondary) | -1.6% | [-4.4%, -0.5%] | 11 |
79+
| All ❌✅ (primary) | -2.8% | [-26.0%, 0.5%] | 13 |
80+
81+
#### Mixed
82+
83+
Enable `CrateNum` query feeding via `TyCtxt` [#123126](https://github.com/rust-lang/rust/pull/123126) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=40dcd796d094b911b8b7b55a0519fb8e3d21680f&end=244da22fabd9fa677bbd0ac601a88e5ca6917526&stat=instructions:u)
84+
85+
| (instructions:u) | mean | range | count |
86+
|:----------------------------------:|:-----:|:--------------:|:-----:|
87+
| Regressions ❌ <br /> (primary) | 0.3% | [0.2%, 0.6%] | 19 |
88+
| Regressions ❌ <br /> (secondary) | 0.4% | [0.3%, 0.5%] | 5 |
89+
| Improvements ✅ <br /> (primary) | - | - | 0 |
90+
| Improvements ✅ <br /> (secondary) | -0.7% | [-0.7%, -0.7%] | 2 |
91+
| All ❌✅ (primary) | 0.3% | [0.2%, 0.6%] | 19 |
92+
93+
This looks like it's mostly just a regression to incremental. The PR
94+
description notes this is expected and sounds like there's follow-up work
95+
planned (unclear whether it will help with performance though).
96+
97+
Stop using LLVM struct types for alloca [#122053](https://github.com/rust-lang/rust/pull/122053) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=c1feb3eceef7d5f0126c309a87062cf413fe0a25&end=29a56a3b1c06a624c0c06728c0af756d09df6b1b&stat=instructions:u)
98+
99+
| (instructions:u) | mean | range | count |
100+
|:----------------------------------:|:-----:|:--------------:|:-----:|
101+
| Regressions ❌ <br /> (primary) | 0.2% | [0.2%, 0.3%] | 8 |
102+
| Regressions ❌ <br /> (secondary) | 0.4% | [0.2%, 1.1%] | 17 |
103+
| Improvements ✅ <br /> (primary) | -1.9% | [-1.9%, -1.9%] | 1 |
104+
| Improvements ✅ <br /> (secondary) | -0.4% | [-0.4%, -0.4%] | 1 |
105+
| All ❌✅ (primary) | -0.0% | [-1.9%, 0.3%] | 9 |
106+
107+
Instruction counts are predominantly affected by some shuffling inside LLVM,
108+
but cycles are largely unaffected. See detailed analysis in
109+
https://github.com/rust-lang/rust/pull/122053#issuecomment-2074850501.
110+
111+
Abort a process when FD ownership is violated [#124210](https://github.com/rust-lang/rust/pull/124210) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=6c90ac8d8f4489472720fce03c338cd5d0977f33&end=cb4940645775f60d74aee2e018d6c516c5aa9ed7&stat=instructions:u)
112+
113+
| (instructions:u) | mean | range | count |
114+
|:----------------------------------:|:-----:|:--------------:|:-----:|
115+
| Regressions ❌ <br /> (primary) | 0.4% | [0.4%, 0.4%] | 2 |
116+
| Regressions ❌ <br /> (secondary) | - | - | 0 |
117+
| Improvements ✅ <br /> (primary) | -0.8% | [-0.8%, -0.8%] | 1 |
118+
| Improvements ✅ <br /> (secondary) | -0.1% | [-0.1%, -0.1%] | 2 |
119+
| All ❌✅ (primary) | 0.0% | [-0.8%, 0.4%] | 3 |
120+
121+
Based on the [self profile results](https://perf.rust-lang.org/detailed-query.html?commit=cb4940645775f60d74aee2e018d6c516c5aa9ed7&benchmark=cargo-0.60.0-opt&scenario=full&base_commit=6c90ac8d8f4489472720fce03c338cd5d0977f33&sort_idx=-12)
122+
I suspect this is caused by us generating more code in the downstream crate(s)
123+
as a result of the late-bound ub checks.

0 commit comments

Comments
 (0)