Add triage-2020-12-15

Author: rylev

triage/2020-12-15.md

@@ -0,0 +1,65 @@
+2020-12-15 Triage Log
+
+A week dominated by small regressions with only 1 modest yet clear performance gain. None of the regressions are large enough to cause concern, but there should be a followup to ensure that those regressions are at least examined.
+
+Triage done by **@rylevick**.
+Revision range: [4fd4a98d4788bc987d7f7add9df5f5ead6a1c15e..e1cce06e4ff5206daf397e1dcf91ed53653be171](https://perf.rust-lang.org/?start=4fd4a98d4788bc987d7f7add9df5f5ead6a1c15e&end=e1cce06e4ff5206daf397e1dcf91ed53653be171&absolute=false&stat=instructions%3Au)
+
+6 Regressions, 1 Improvements, 2 Mixed
+0 of them in rollups
+
+#### Regressions
+
+Also generate `StorageDead` in constants[#78679](https://github.com/rust-lang/rust/issues/78679)
+- Large regression in [instruction counts](https://perf.rust-lang.org/compare.html?start=c0bfe3485f97f267cc8adec724f109c56dab5526&end=cc03ee6702053ded253c3656cbd02f0bfdf25c73&stat=instructions:u) (up to 5.7% on `incr-patched: new row` builds of `tuple-stress-check`)
+- A removal of special casing of not marking statics and constants as `StorageDead` inside rustc_mir. 
+- This was regression was [deemed as acceptable](https://github.com/rust-lang/rust/pull/78679#issuecomment-729030782) due to being more correct than the previous implementation. 
+
+Properly re-use def path hash in incremental mode[#79721](https://github.com/rust-lang/rust/issues/79721)
+- Moderate regression in [instruction counts](https://perf.rust-lang.org/compare.html?start=cc03ee6702053ded253c3656cbd02f0bfdf25c73&end=fa55f668e5ea5388ec98b9340969527252239151&stat=instructions:u) (up to 1.5% on `incr-unchanged` builds of `deeply-nested-opt`)
+- Fixes a p-high [ICE issue](https://github.com/rust-lang/rust/issues/79661), and so the perf regressions were deemed acceptable.
+
+Accept arbitrary expressions in key-value attributes at parse time[#78837](https://github.com/rust-lang/rust/issues/78837)
+- Moderate regression in [instruction counts](https://perf.rust-lang.org/compare.html?start=1cc410710993d036730c11556039e40109f6ab41&end=58d2bad9f7ab0971495247b6c94978848760ca9d&stat=instructions:u) (up to 1.4% on `incr-unchanged` builds of `match-stress-exhaustive_patterns-check`)
+- [Pinged in the PR about this issue](https://github.com/rust-lang/rust/pull/78837#issuecomment-745380762)
+
+Capture precise paths in THIR and MIR[#79553](https://github.com/rust-lang/rust/issues/79553)
+- Moderate regression in [instruction counts](https://perf.rust-lang.org/compare.html?start=9eb3a7ceafd1e2c1924177caa18c7cc0c25b413e&end=5bd9b60333b3dc0a51e7a5607cd1e0d537a9f718&stat=instructions:u) (up to 4.5% on `incr-unchanged` builds of `clap-rs-check`)
+- While this change powers a feature behind a feature flag (`capture_disjoint_fields`), it looks like it's still causing perf regressions in workloads not using this feature.
+- [Pinged in the PR about this issue](https://github.com/rust-lang/rust/pull/79553#issuecomment-745435558)
+
+Create `rustc_type_ir`[#79169](https://github.com/rust-lang/rust/issues/79169)
+- Moderate regression in [instruction counts](https://perf.rust-lang.org/compare.html?start=c3ed6681ff8d446e68ce272be4bf66f4145f6e29&end=3f2088aa603d2cd3f43c20795872de9cd6ec7735&stat=instructions:u) (up to 3.1% on `full` builds of `ctfe-stress-4-check`)
+* This mainly seems to be moving code around so it might be an inlining issue. 
+* [Pinged in the PR about this issue](https://github.com/rust-lang/rust/pull/79169#issuecomment-745388674)
+
+Update stdarch submodule[#79938](https://github.com/rust-lang/rust/issues/79938)
+- Moderate regression in [instruction counts](https://perf.rust-lang.org/compare.html?start=fa416394275d2468d104b8f72ac31b1ddf7ee52e&end=8b3ee82eb68cb35030bb745c23f8aa76d9de5bee&stat=instructions:u) (up to 1.4% on `incr-unchanged` builds of `deeply-nested-debug`)
+- This was a wholesale update of the stdarch submodule. 
+* [Pinged in the PR about this issue](https://github.com/rust-lang/rust/pull/79938#issuecomment-745393740)
+
+#### Improvements
+
+Compress RWU from at least 32 bits to 4 bits[#79727](https://github.com/rust-lang/rust/issues/79727)
+- Moderate improvement in [instruction counts](https://perf.rust-lang.org/compare.html?start=5e6e1e33a11d140a4d70f946730137f241224eb3&end=1700ca07c6dd7becff85678409a5df6ad4cf4f47&stat=instructions:u) (up to -4.8% on `full` builds of `clap-rs-check`)
+- This was explicitly an experiment to gain performance, and it seems to worked fairly well. Other bit representations were tested but the one chosen was the most efficient. 
+
+#### Mixed
+
+Use `def_path_hash_to_def_id` when re-using a `RawDefId`[#79915](https://github.com/rust-lang/rust/issues/79915)
+- Moderate improvement in [instruction counts](https://perf.rust-lang.org/compare.html?start=a2e29d67c26bdf8f278c98ee02d6cc77a279ed2e&end=19eb1c4c526071c430c05fffc64da71ac057a3d5&stat=instructions:u) (up to -3.5% on `incr-unchanged` builds of `clap-rs-check`)
+- Moderate regression in [instruction counts](https://perf.rust-lang.org/compare.html?start=a2e29d67c26bdf8f278c98ee02d6cc77a279ed2e&end=19eb1c4c526071c430c05fffc64da71ac057a3d5&stat=instructions:u) (up to 2.7% on `incr-patched: dummy fn` builds of `unused-warnings-check`)
+- The amount of regressions outweighs the improvements (which were just in the clap benchmark).
+- This is a followup fix to [#79721](https://github.com/rust-lang/rust/issues/79721) which also introduced perf regressions.
+- [Pinged on the issue](https://github.com/rust-lang/rust/pull/79915#issuecomment-745409364)
+
+Lower `discriminant_value` intrinsic[#79922](https://github.com/rust-lang/rust/issues/79922)
+- Moderate improvement in [instruction counts](https://perf.rust-lang.org/compare.html?start=8b3ee82eb68cb35030bb745c23f8aa76d9de5bee&end=5d77fc8d0db3b69f3a3691d86eba23e4cdc390e1&stat=instructions:u) (up to -3.9% on `full` builds of `match-stress-enum-check`)
+- Smaller regression in [instruction counts](https://perf.rust-lang.org/compare.html?start=8b3ee82eb68cb35030bb745c23f8aa76d9de5bee&end=5d77fc8d0db3b69f3a3691d86eba23e4cdc390e1&stat=instructions:u) (up to 1.3% on `incr-unchanged` builds of `clap-rs-check`)
+- The improvement outweighs the regression.
+
+#### Nags requiring follow up
+
+- Several regressions need followup investigations. See their respective entries above for the issue.
+- As mentioned last week, stdarch expansion causing a 40% libcore compile time regression is still not
+  resolved, and resolution is unclear.