|
| 1 | +# 2022-10-11 Triage Log |
| 2 | + |
| 3 | +Overall, a fairly quiet week where the change to primary benchmarks ended up breaking exactly even. Secondary benchmarks saw improvements but not in large enough numbers for it to be particularly noteworthy. |
| 4 | + |
| 5 | +Triage done by **@rylev**. |
| 6 | +Revision range: [02cd79a..1e926f0](https://perf.rust-lang.org/?start=02cd79afb8080fce8c8ce35533c54d8ecf8f390e&end=1e926f06528ecb2503f026e2fd53cb735d487b10&absolute=false&stat=instructions%3Au) |
| 7 | + |
| 8 | +**Summary**: |
| 9 | + |
| 10 | +| (instructions:u) | mean | range | count | |
| 11 | +|:----------------:|:----:|:-----:|:-----:| |
| 12 | +| Regressions ❌ <br /> (primary) | 0.8% | [0.2%, 1.4%] | 19 | |
| 13 | +| Regressions ❌ <br /> (secondary) | 1.0% | [0.3%, 1.8%] | 9 | |
| 14 | +| Improvements ✅ <br /> (primary) | -0.6% | [-1.8%, -0.3%] | 29 | |
| 15 | +| Improvements ✅ <br /> (secondary) | -1.0% | [-6.4%, -0.2%] | 39 | |
| 16 | +| All ❌✅ (primary) | -0.0% | [-1.8%, 1.4%] | 48 | |
| 17 | + |
| 18 | + |
| 19 | +3 Regressions, 1 Improvements, 6 Mixed; 4 of them in rollups |
| 20 | +41 artifact comparisons made in total |
| 21 | + |
| 22 | +#### Regressions |
| 23 | + |
| 24 | +Reduce CString allocations in std as much as possible [#93668](https://github.com/rust-lang/rust/pull/93668) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=79a664d8b00505a76b53cfe017b9c80bcee7e080&end=1b225414f325593f974c6b41e671a0a0dc5d7d5e&stat=instructions:u) |
| 25 | + |
| 26 | +| (instructions:u) | mean | range | count | |
| 27 | +|:----------------:|:----:|:-----:|:-----:| |
| 28 | +| Regressions ❌ <br /> (primary) | 1.0% | [1.0%, 1.0%] | 4 | |
| 29 | +| Regressions ❌ <br /> (secondary) | 0.2% | [0.2%, 0.2%] | 2 | |
| 30 | +| Improvements ✅ <br /> (primary) | - | - | 0 | |
| 31 | +| Improvements ✅ <br /> (secondary) | - | - | 0 | |
| 32 | +| All ❌✅ (primary) | 1.0% | [1.0%, 1.0%] | 4 | |
| 33 | +- The `hello-world` opt benchmarks are dominated by link time. |
| 34 | +- It makes sense that a change to an FFI type `CString` could have an impact on these. |
| 35 | +- I don't think there's a need though to really do anything about it. |
| 36 | + |
| 37 | + |
| 38 | +Rollup of 6 pull requests [#102867](https://github.com/rust-lang/rust/pull/102867) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=1a7c203e7f89f92598a793c6f6406735c024e7ee&end=e495b37c9a301d776a7bd0c72d5c4a202e159032&stat=instructions:u) |
| 39 | + |
| 40 | +| (instructions:u) | mean | range | count | |
| 41 | +|:----------------:|:----:|:-----:|:-----:| |
| 42 | +| Regressions ❌ <br /> (primary) | 0.3% | [0.3%, 0.4%] | 6 | |
| 43 | +| Regressions ❌ <br /> (secondary) | 1.4% | [1.2%, 1.6%] | 6 | |
| 44 | +| Improvements ✅ <br /> (primary) | - | - | 0 | |
| 45 | +| Improvements ✅ <br /> (secondary) | -0.2% | [-0.2%, -0.2%] | 1 | |
| 46 | +| All ❌✅ (primary) | 0.3% | [0.3%, 0.4%] | 6 | |
| 47 | +- The impacted benchmarks are more sensitive to changes to the trait system, so it looks like it might be [#102845](https://github.com/rust-lang/rust/pull/102845). |
| 48 | +- Kicked of a perf run to investigate. |
| 49 | + |
| 50 | + |
| 51 | +tools/remote-test-{server,client}: Use /data/local/tmp on Android [#102755](https://github.com/rust-lang/rust/pull/102755) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=365578445c6771bb74bfd341c8c48867164a9992&end=1e926f06528ecb2503f026e2fd53cb735d487b10&stat=instructions:u) |
| 52 | + |
| 53 | +| (instructions:u) | mean | range | count | |
| 54 | +|:----------------:|:----:|:-----:|:-----:| |
| 55 | +| Regressions ❌ <br /> (primary) | 0.7% | [0.6%, 0.9%] | 6 | |
| 56 | +| Regressions ❌ <br /> (secondary) | - | - | 0 | |
| 57 | +| Improvements ✅ <br /> (primary) | - | - | 0 | |
| 58 | +| Improvements ✅ <br /> (secondary) | - | - | 0 | |
| 59 | +| All ❌✅ (primary) | 0.7% | [0.6%, 0.9%] | 6 | |
| 60 | +- Looks like Diesel is becoming more noisy lately. You can see that in [this graph](https://user-images.githubusercontent.com/1327285/195131717-d78400d3-218b-4bb2-967a-7faef12b9fbe.png). |
| 61 | + |
| 62 | + |
| 63 | +#### Improvements |
| 64 | + |
| 65 | +Rewrite representability [#100720](https://github.com/rust-lang/rust/pull/100720) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=a688a0305fad9219505a8f2576446510601bafe8&end=bba9785dd73f61aacd301a2cb379e1e85a129047&stat=instructions:u) |
| 66 | + |
| 67 | +| (instructions:u) | mean | range | count | |
| 68 | +|:----------------:|:----:|:-----:|:-----:| |
| 69 | +| Regressions ❌ <br /> (primary) | - | - | 0 | |
| 70 | +| Regressions ❌ <br /> (secondary) | - | - | 0 | |
| 71 | +| Improvements ✅ <br /> (primary) | -0.4% | [-0.8%, -0.2%] | 38 | |
| 72 | +| Improvements ✅ <br /> (secondary) | -0.9% | [-3.3%, -0.2%] | 21 | |
| 73 | +| All ❌✅ (primary) | -0.4% | [-0.8%, -0.2%] | 38 | |
| 74 | + |
| 75 | + |
| 76 | +#### Mixed |
| 77 | + |
| 78 | +Remove `TypeckResults` from `InferCtxt` [#101632](https://github.com/rust-lang/rust/pull/101632) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=e42c4d7218b2596276152c5eb1e69335621f3086&end=43c22af267fd9337bc05382b2771dde49d2e9f26&stat=instructions:u) |
| 79 | + |
| 80 | +| (instructions:u) | mean | range | count | |
| 81 | +|:----------------:|:----:|:-----:|:-----:| |
| 82 | +| Regressions ❌ <br /> (primary) | 0.7% | [0.5%, 1.2%] | 13 | |
| 83 | +| Regressions ❌ <br /> (secondary) | 4.3% | [3.2%, 5.7%] | 6 | |
| 84 | +| Improvements ✅ <br /> (primary) | -0.3% | [-0.6%, -0.2%] | 19 | |
| 85 | +| Improvements ✅ <br /> (secondary) | -0.6% | [-1.6%, -0.2%] | 52 | |
| 86 | +| All ❌✅ (primary) | 0.1% | [-0.6%, 1.2%] | 32 | |
| 87 | +- Looks `specialization_graph::Children::insert` is getting called way more. |
| 88 | +- Perhaps some strategic placement of `inline` could help help. |
| 89 | + |
| 90 | + |
| 91 | +Rollup of 6 pull requests [#102787](https://github.com/rust-lang/rust/pull/102787) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=43c22af267fd9337bc05382b2771dde49d2e9f26&end=2d3a85b4f8ba7e2554f4d4fee126bc2ac6ee2af4&stat=instructions:u) |
| 92 | + |
| 93 | +| (instructions:u) | mean | range | count | |
| 94 | +|:----------------:|:----:|:-----:|:-----:| |
| 95 | +| Regressions ❌ <br /> (primary) | 0.3% | [0.2%, 0.3%] | 2 | |
| 96 | +| Regressions ❌ <br /> (secondary) | 1.4% | [1.1%, 1.9%] | 6 | |
| 97 | +| Improvements ✅ <br /> (primary) | -0.8% | [-1.0%, -0.4%] | 8 | |
| 98 | +| Improvements ✅ <br /> (secondary) | -2.5% | [-3.7%, -0.3%] | 7 | |
| 99 | +| All ❌✅ (primary) | -0.6% | [-1.0%, 0.3%] | 10 | |
| 100 | +- Most of the regressions are in secondary benchmarks, so I don't think it's worth investigating what caused this. |
| 101 | + |
| 102 | + |
| 103 | +std: use futex in `Once` [#99505](https://github.com/rust-lang/rust/pull/99505) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=eed7f2f58bd44d32ac30e48425eb4bede7ea84f7&end=a688a0305fad9219505a8f2576446510601bafe8&stat=instructions:u) |
| 104 | + |
| 105 | +| (instructions:u) | mean | range | count | |
| 106 | +|:----------------:|:----:|:-----:|:-----:| |
| 107 | +| Regressions ❌ <br /> (primary) | 0.5% | [0.5%, 0.5%] | 1 | |
| 108 | +| Regressions ❌ <br /> (secondary) | 1.7% | [1.0%, 3.3%] | 7 | |
| 109 | +| Improvements ✅ <br /> (primary) | - | - | 0 | |
| 110 | +| Improvements ✅ <br /> (secondary) | -0.3% | [-0.5%, -0.2%] | 9 | |
| 111 | +| All ❌✅ (primary) | 0.5% | [0.5%, 0.5%] | 1 | |
| 112 | +- The regression results are small and neutral enough that we don't need to investigate. |
| 113 | + |
| 114 | + |
| 115 | +Rollup of 8 pull requests [#102809](https://github.com/rust-lang/rust/pull/102809) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=bba9785dd73f61aacd301a2cb379e1e85a129047&end=c27948d255167d827421401950c6d723ba28de8f&stat=instructions:u) |
| 116 | + |
| 117 | +| (instructions:u) | mean | range | count | |
| 118 | +|:----------------:|:----:|:-----:|:-----:| |
| 119 | +| Regressions ❌ <br /> (primary) | 0.3% | [0.2%, 0.4%] | 13 | |
| 120 | +| Regressions ❌ <br /> (secondary) | 0.4% | [0.3%, 0.6%] | 3 | |
| 121 | +| Improvements ✅ <br /> (primary) | - | - | 0 | |
| 122 | +| Improvements ✅ <br /> (secondary) | -1.1% | [-1.1%, -1.1%] | 1 | |
| 123 | +| All ❌✅ (primary) | 0.3% | [0.2%, 0.4%] | 13 | |
| 124 | + |
| 125 | + |
| 126 | +Rollup of 6 pull requests [#102875](https://github.com/rust-lang/rust/pull/102875) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=8dfb40722da197e77a33a19eb9d3fd6512831341&end=691aeaad6c9ee3483f2851695379b348b8a92a43&stat=instructions:u) |
| 127 | + |
| 128 | +| (instructions:u) | mean | range | count | |
| 129 | +|:----------------:|:----:|:-----:|:-----:| |
| 130 | +| Regressions ❌ <br /> (primary) | 0.2% | [0.2%, 0.2%] | 2 | |
| 131 | +| Regressions ❌ <br /> (secondary) | - | - | 0 | |
| 132 | +| Improvements ✅ <br /> (primary) | - | - | 0 | |
| 133 | +| Improvements ✅ <br /> (secondary) | -5.0% | [-6.6%, -1.8%] | 5 | |
| 134 | +| All ❌✅ (primary) | 0.2% | [0.2%, 0.2%] | 2 | |
| 135 | +- This is neutral enough that I don't believe it warrants investigation. |
| 136 | + |
| 137 | + |
| 138 | +slice: #[inline] a couple iterator methods. [#96711](https://github.com/rust-lang/rust/pull/96711) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=691aeaad6c9ee3483f2851695379b348b8a92a43&end=0265a3e93bf1b89d97cae113ed214954d5c35e22&stat=instructions:u) |
| 139 | + |
| 140 | +| (instructions:u) | mean | range | count | |
| 141 | +|:----------------:|:----:|:-----:|:-----:| |
| 142 | +| Regressions ❌ <br /> (primary) | - | - | 0 | |
| 143 | +| Regressions ❌ <br /> (secondary) | 0.5% | [0.3%, 0.7%] | 4 | |
| 144 | +| Improvements ✅ <br /> (primary) | -0.8% | [-1.5%, -0.5%] | 8 | |
| 145 | +| Improvements ✅ <br /> (secondary) | -1.4% | [-1.8%, -1.2%] | 6 | |
| 146 | +| All ❌✅ (primary) | -0.8% | [-1.5%, -0.5%] | 8 | |
| 147 | +- From the reviewer: "Perf results are more positive than negative, I think that's all that matters for this kind of change. The regressions are minor ones in secondary benchmarks" |
| 148 | + |
0 commit comments