Repository navigation
Missing coverage on assert!(!...) statements #118904
Description
I tried this code:
fn main() {
assert!(!false);
}I expected to see this happen: Run this program with cargo-llvm-cov: cargo llvm-cov run --html, I expect the source code should be fully covered.
Instead, this happened: The assert! part in the source code is not covered. But if I change !false to true, the source will be fully covered, it seems that I can’t get assert! to be covered as long as the argument starts with an ! operator.
Meta
rustc --version --verbose:
rustc 1.76.0-nightly (f967532a4 2023-12-08)
binary: rustc
commit-hash: f967532a47eb728ada44473a5c4c2eca1a45fe30
commit-date: 2023-12-08
host: x86_64-pc-windows-msvc
release: 1.76.0-nightly
LLVM version: 17.0.5
Activity
[-]Missing coverage on expression `assert!(!...)` statement[/-][+]Missing coverage on `assert!(!...)` statements[/+]matthiaskrgr commentedon Dec 13, 2023
Might be that llvm just optimizes it out as dead code. Do you see it with -Copt-level=0 ?
EFanZh commentedon Dec 13, 2023
@matthiaskrgr: I think
cargo-llvm-covalready usesopt-level=0by default. Nevertheless, I have tested usingenv RUSTFLAGS='-Copt-level=0' cargo llvm-cov run --html, and the result is the same.taiki-e commentedon Dec 13, 2023
This may be a similar case to #110486. It seems that something odd is happening when the compiler gets the coverage of the code generated by the macro and maps it to the original code.
@rustbot label +A-code-coverage
Noratrieb commentedon Dec 15, 2023
cc @Zalathar
Zalathar commentedon Dec 16, 2023
I can reproduce this with
2023-12-08but not with2023-12-14.It was probably fixed by #118198.
Noratrieb commentedon Dec 16, 2023
Would it make sense to add a test for this specifically?
Zalathar commentedon Dec 16, 2023
PR filed: #119011
Rollup merge of rust-lang#119011 - Zalathar:assert-not, r=est31
Auto merge of rust-lang#119011 - Zalathar:assert-not, r=est31
ranger-ross commentedon Aug 31, 2024
A bit late, but I re-tested this on Rust 1.80.0 and it look like #119011 fixed the issue.
I think we are good to close this issue 👍