David Chase

-obj did not do what the doc lead me to believe it should

7

I typed: `compilecmp -all -obj` I expected to see some object file sizes, but instead I saw this: ``` compilecmp master HEAD 14:00:39 copy tree at master (90b39f33c40e8d4ae085426313be69473f304e7d) to /usr/local/google/home/drchase/.compilecmp/90b39f33c40e8d4ae085426313be69473f304e7d...

Not sure if this counts as a bug (I am torturing fzgen in ways you might not have imagined -- I [automated it against the benchmarks in the "bent" collection](https://go-review.googlesource.com/c/benchmarks/+/373504)),...

Instructions in ReadMe.md are ambiguous or out-of-date.

1

The quick start instructions in ReadMe.md are either ambiguous or out of date; i.e., I attempted to follow them, but could not get it to run. It appears to want...

Drag scrolling (on macOS laptop with a retina display) is too fast

4

There may be a way to adjust this, but I could not find it. What I want is to scroll with drag gestures on my trackpad; this "works" but instead...

update unsafe code to user newer-faster-better idioms

1

The new unsafe string/slice functions insulate the code from unlikely-but-still-possible changes to string and slice representation, and tend to compile to slightly better code. Build tags keep things working with...

Gollvm inlines aggressively and discovers more dead code, resulting in very bogus benchmark results (zero nanoseconds). This prevents this (for this benchmark; others may need the same treatment).

structs: add HostLayout "directive" type

47

### Proposal Details ### Abstract This proposes a new package for zero-sized types whose presence in a structure’s list of fields would control how the compiler lays out those fields,...

Representative benchmarks of wazero for Go compiler perf measurement

9

**Is your feature request related to a problem? Please describe.** I'd like to know which of the wazero benchmarks (specific Benchmark... functions) are most representative of wazero user code, if...