Release Notes for 1.59 Missing Change to profile.dev optimization

[TheBlueMatt]

Problem

On 1.58, [profile.dev] opt-level=1 implies that a crate being tested with cargo test without a -C opt-level argument, but builds all dependent crates (including workspace crates, if there's a workspace dep, from what I can see) with -C opt-level=1. On 1.59 this changed to include all crates getting compiled with -C opt-level=1, increasing compile time and increasing performance (obviously).

Given its been a while, it seems like the correct solution is to simply amend the release notes to point this out.

CC rust-lang/rust#97460 (comment)

Steps

No response

Possible Solution(s)

No response

Notes

No response

Version

No response

[ehuss]

Hm, I don't believe there were any changes to profiles in 1.59. Unfortunately, I can't reproduce what you are referring to. Can you put together a set of steps to reproduce the issue? For example:

cargo new foo
cd foo

cat >> Cargo.toml << EOF
[dependencies]
itoa = "1.0"
[profile.dev]
opt-level = 1
EOF

cargo +1.58.0 test -v

should produce output similar to:

   Compiling itoa v1.0.2
     Running `rustc --crate-name itoa … -C opt-level=1 …
   Compiling foo v0.1.0 (/Users/eric/Temp/foo)
     Running `rustc --crate-name foo ¬ -C opt-level=1 … --test …
    Finished test [optimized + debuginfo] target(s) in 1.03s
     Running `/Users/eric/Temp/foo/target/debug/deps/foo-d949ffb9e23fe5ab`

running 0 tests

test result: ok. 0 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s

[TheBlueMatt]

Ah, I'm so sorry, was looking at the wrong thing while debugging rust-lang/rust#97460, this isn't a bug, its just the namespace change from 1.57.