rust-lang · Kobzol · Nov 26, 2024
diff --git a/triage/2024-11-26.md b/triage/2024-11-26.md
@@ -0,0 +1,149 @@
+# 2024-11-26 Triage Log
+
+This week saw more regressions than improvements, mostly due to three PRs that performed internal
+refactorings that are necessary for further development and modification of the compiler.
+
+Triage done by **@kobzol**.
+Revision range: [7d40450b..7db7489f](https://perf.rust-lang.org/?start=7d40450b2df92bdc9dec414b30cf5f7a5979a92e&end=7db7489f9bc274cb60c4956bfa56de0185eb1b9b&absolute=false&stat=instructions%3Au)
+
+**Summary**:
+
+| (instructions:u)                   | mean  | range          | count |
+|:----------------------------------:|:-----:|:--------------:|:-----:|
+| Regressions ❌ <br /> (primary)    | 0.6%  | [0.1%, 3.6%]   | 57    |
+| Regressions ❌ <br /> (secondary)  | 0.6%  | [0.0%, 2.7%]   | 100   |
+| Improvements ✅ <br /> (primary)   | -0.5% | [-1.5%, -0.2%] | 11    |
+| Improvements ✅ <br /> (secondary) | -0.4% | [-0.5%, -0.3%] | 7     |
+| All ❌✅ (primary)                 | 0.4%  | [-1.5%, 3.6%]  | 68    |
+
+
+4 Regressions, 2 Improvements, 3 Mixed; 3 of them in rollups
+40 artifact comparisons made in total
+
+#### Regressions
+
+Rollup of 8 pull requests [#133219](https://github.com/rust-lang/rust/pull/133219) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=ee612c45f00391aff71ec0c52b7fc08fae18c711&end=875df370be806c837f58abb638329905e969ace4&stat=instructions:u)
+
+| (instructions:u)                   | mean | range        | count |
+|:----------------------------------:|:----:|:------------:|:-----:|
+| Regressions ❌ <br /> (primary)    | 0.7% | [0.2%, 1.1%] | 6     |
+| Regressions ❌ <br /> (secondary)  | 0.5% | [0.3%, 1.2%] | 13    |
+| Improvements ✅ <br /> (primary)   | -    | -            | 0     |
+| Improvements ✅ <br /> (secondary) | -    | -            | 0     |
+| All ❌✅ (primary)                 | 0.7% | [0.2%, 1.1%] | 6     |
+
+- The regressions were on the helloworld benchmark and other tiny crates, and they were reverted
+in a follow-up merge commit.
+- Marked as triaged.
+
+continue `ParamEnv` to `TypingEnv` transition [#133212](https://github.com/rust-lang/rust/pull/133212) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=bcfea1f8d253fd43eab36832fa601d192ac603d5&end=70e814bd9e532a302763f870c665c5af59c2b632&stat=instructions:u)
+
+| (instructions:u)                   | mean | range        | count |
+|:----------------------------------:|:----:|:------------:|:-----:|
+| Regressions ❌ <br /> (primary)    | 0.4% | [0.3%, 0.5%] | 5     |
+| Regressions ❌ <br /> (secondary)  | 2.1% | [0.1%, 6.6%] | 12    |
+| Improvements ✅ <br /> (primary)   | -    | -            | 0     |
+| Improvements ✅ <br /> (secondary) | -    | -            | 0     |
+| All ❌✅ (primary)                 | 0.4% | [0.3%, 0.5%] | 5     |
+
+- Regressions caused by internal refactoring of the type system that is necessary for further
+improvements.
+- Asked the PR author about the perf. effect of this PR and also
+of [#133242](https://github.com/rust-lang/rust/pull/133242) (mentioned below).
+- Not marking as triaged.
+
+Rollup of 5 pull requests [#133280](https://github.com/rust-lang/rust/pull/133280) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=318f96a8cf3eca5c4aaf60a992f349bce5c3fd41&end=0b1bf71a71c2a1d34c212285362530ec2c4e4775&stat=instructions:u)
+
+| (instructions:u)                   | mean | range        | count |
+|:----------------------------------:|:----:|:------------:|:-----:|
+| Regressions ❌ <br /> (primary)    | 0.3% | [0.2%, 0.4%] | 7     |
+| Regressions ❌ <br /> (secondary)  | 2.0% | [2.0%, 2.0%] | 1     |
+| Improvements ✅ <br /> (primary)   | -    | -            | 0     |
+| Improvements ✅ <br /> (secondary) | -    | -            | 0     |
+| All ❌✅ (primary)                 | 0.3% | [0.2%, 0.4%] | 7     |
+
+- The secondary regression is noise.
+- The primary regressions look to be genuine, but they are only on a single benchmark and they are
+very small, so I don't think that further investigation is needed.
+- Marked as triaged.
+
+finish `Reveal` removal [#133242](https://github.com/rust-lang/rust/pull/133242) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=826b673412d2f579e7865b6f26bc9771c6d3b097&end=386a7c7ae22a3727c6a06633212d5a95edabdd74&stat=instructions:u)
+
+| (instructions:u)                   | mean | range        | count |
+|:----------------------------------:|:----:|:------------:|:-----:|
+| Regressions ❌ <br /> (primary)    | 0.7% | [0.2%, 3.6%] | 11    |
+| Regressions ❌ <br /> (secondary)  | 0.3% | [0.2%, 0.4%] | 24    |
+| Improvements ✅ <br /> (primary)   | -    | -            | 0     |
+| Improvements ✅ <br /> (secondary) | -    | -            | 0     |
+| All ❌✅ (primary)                 | 0.7% | [0.2%, 3.6%] | 11    |
+
+- Similar situation as in [#133212](https://github.com/rust-lang/rust/pull/133212).
+- Not marking as triaged.
+
+#### Improvements
+
+Resolve tweaks [#132761](https://github.com/rust-lang/rust/pull/132761) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=78993684f2ba22082197d63627f3e91069911839&end=ee612c45f00391aff71ec0c52b7fc08fae18c711&stat=instructions:u)
+
+| (instructions:u)                   | mean  | range          | count |
+|:----------------------------------:|:-----:|:--------------:|:-----:|
+| Regressions ❌ <br /> (primary)    | -     | -              | 0     |
+| Regressions ❌ <br /> (secondary)  | -     | -              | 0     |
+| Improvements ✅ <br /> (primary)   | -0.3% | [-0.6%, -0.1%] | 68    |
+| Improvements ✅ <br /> (secondary) | -1.1% | [-1.1%, -1.1%] | 1     |
+| All ❌✅ (primary)                 | -0.3% | [-0.6%, -0.1%] | 68    |
+
+
+Drop debug info instead of panicking if we exceed LLVM's capability to represent it [#133194](https://github.com/rust-lang/rust/pull/133194) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=875df370be806c837f58abb638329905e969ace4&end=bcfea1f8d253fd43eab36832fa601d192ac603d5&stat=instructions:u)
+
+| (instructions:u)                   | mean  | range          | count |
+|:----------------------------------:|:-----:|:--------------:|:-----:|
+| Regressions ❌ <br /> (primary)    | -     | -              | 0     |
+| Regressions ❌ <br /> (secondary)  | -     | -              | 0     |
+| Improvements ✅ <br /> (primary)   | -1.1% | [-1.1%, -1.1%] | 2     |
+| Improvements ✅ <br /> (secondary) | -0.4% | [-0.7%, -0.4%] | 8     |
+| All ❌✅ (primary)                 | -1.1% | [-1.1%, -1.1%] | 2     |
+
+
+#### Mixed
+
+Rollup of 5 pull requests [#133234](https://github.com/rust-lang/rust/pull/133234) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=70e814bd9e532a302763f870c665c5af59c2b632&end=fda68927475070696fcc9d1f5c9c990f0e1af87a&stat=instructions:u)
+
+| (instructions:u)                   | mean  | range          | count |
+|:----------------------------------:|:-----:|:--------------:|:-----:|
+| Regressions ❌ <br /> (primary)    | 2.2%  | [0.2%, 5.5%]   | 8     |
+| Regressions ❌ <br /> (secondary)  | 0.8%  | [0.8%, 0.8%]   | 1     |
+| Improvements ✅ <br /> (primary)   | -     | -              | 0     |
+| Improvements ✅ <br /> (secondary) | -5.9% | [-5.9%, -5.9%] | 1     |
+| All ❌✅ (primary)                 | 2.2%  | [0.2%, 5.5%]   | 8     |
+
+- The regression was caused by [#132732](https://github.com/rust-lang/rust/pull/132732).
+- Asked the PR author if there is a way to reduce the compilation performance cost.
+- Not marking as triaged.
+
+Reduce false positives of tail-expr-drop-order from consumed values (attempt #2) [#131326](https://github.com/rust-lang/rust/pull/131326) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=a1f299953656f95004c69b24ad8071d6899fa9da&end=3fee0f12e4f595948f8f54f57c8b7a7a58127124&stat=instructions:u)
+
+| (instructions:u)                   | mean  | range          | count |
+|:----------------------------------:|:-----:|:--------------:|:-----:|
+| Regressions ❌ <br /> (primary)    | 0.2%  | [0.2%, 0.2%]   | 1     |
+| Regressions ❌ <br /> (secondary)  | 0.5%  | [0.2%, 2.3%]   | 15    |
+| Improvements ✅ <br /> (primary)   | -0.7% | [-2.2%, -0.2%] | 19    |
+| Improvements ✅ <br /> (secondary) | -0.6% | [-1.0%, -0.3%] | 8     |
+| All ❌✅ (primary)                 | -0.7% | [-2.2%, 0.2%]  | 20    |
+
+- This was a win on primary benchmarks.
+- Marked as triaged.
+
+#124141 preliminaries [#132629](https://github.com/rust-lang/rust/pull/132629) [(Comparison Link)](https://perf.rust-lang.org/compare.html?start=0b1bf71a71c2a1d34c212285362530ec2c4e4775&end=717f5df2c308dfb4b7b8e6c002c11fe8269c4011&stat=instructions:u)
+
+| (instructions:u)                   | mean  | range          | count |
+|:----------------------------------:|:-----:|:--------------:|:-----:|
+| Regressions ❌ <br /> (primary)    | 0.2%  | [0.1%, 0.3%]   | 29    |
+| Regressions ❌ <br /> (secondary)  | 0.3%  | [0.2%, 0.4%]   | 16    |
+| Improvements ✅ <br /> (primary)   | -     | -              | 0     |
+| Improvements ✅ <br /> (secondary) | -0.4% | [-0.5%, -0.3%] | 5     |
+| All ❌✅ (primary)                 | 0.2%  | [0.1%, 0.3%]   | 29    |
+
+- The regressions are caused by changes in the parser that are required for further cleanup and
+refactoring work.
+- The regressions are genuine, but tiny. The author has deemed them to be unavoidable.
+- Already marked as triaged by @nnethercote.