2020-08-17 Triage Log

See the README for details on how this log was gathered.

This week was quite noisy, with many small regressions and improvements. Overall, instruction counts have increased somewhat, while max RSS remains the same. There were several updates to core data structures (HashMap, IndexMap), which contributed to some of the noise.

Triage done by @ecstaticmorse. Revision range: 8b84156c6ee2b4e707dc32f5a516e3143ab924d3..67e7b9b8cf776222825dbbd4cb1e39b7765ef27c

4 Regressions, 3 Improvements, 4 Mixed, 0 of them in rollups.

4 Regressions

#73851 Remove most specialization use in serialization

• Very small regressions across the board

#75048 Prevent __rust_begin_short_backtrace frames from being tail-call optimised away

• Regressions in several benchmarks
• Mostly for crates with very little codegen

#75306 Hash parent ExpnData

• Very small regressions across the board
• Necessary for correctness

#73656 move Deaggregate pass to post_borrowck_cleanup

• Very small regressions across the board

3 Improvements

#74512 Put panic code path from copy_from_slice into cold function

• Small improvements across the board, albeit mostly for incremental builds

#74877 Implement the min_const_generics feature gate

• A small improvement on wf-projection-stress

#75121 Avoid unwrap_or_else in str indexing

• A small improvement on encoding-opt

4 Mixed

#70052 Update hashbrown to 0.8.1

• Mixed results across the board

#75278 Upgrade indexmap and use it more

• Very small regressions for most builds, notably docs
• Offset by larger wins on a few benchmarks: encoding, keccak, inflate.
• Author cites improved CPU cycle counts.

#75306 Update hashbrown to 0.8.2

• Mixed results across the board
• Mostly improvements, but incr-patched: println has taken a rather large hit.

#75382 First iteration of simplify match branches

• Very small improvements across the board
• However, a few moderate regressions on synthetic benchmarks

Nags requiring follow up

• #73656
• #73851
• #75048