CodSpeed Performance Report

Merging #5595 will not alter performance

_{Comparing tkelman:lief-load-command-types (074df06) with 25.1.x (6873259)}

Summary

✅ 5 untouched benchmarks

Thanks a lot for the PR! Given that this still breaks with 0.15 (and thus the version that conda-build 25.1 claimed compatibility to), I think it would be a good idea to backport this patch.

I see you've got 25.1.x and 24.11.x branches here, but I don't think #5565 was included in 24.11.x. would you like me to open a separate PR, or change the target branch, or will you handle the cherry-picking?

No need for action right now. First off, a backport is not my decision, and secondly, I think it needs to land on main first anyway

I think we can tsrhet 25.1.x. That branch will be synced with main later.

I think we can target 25.1.x.

Could you retarget the branch then? Contributors can't do it, it needs maintainer rights

Edit: apparently, I was wrong. 🤷 🙃

I tried that out but it looked a little messy adding in 4 unrelated commits that exist on main but not the 25.1.x branch

it looked a little messy adding in 4 unrelated commits that exist on main but not the 25.1.x branch

Obviously you'd have to rebase out those commits. git rebase -i 25.1.x and then remove those commits that aren't yours.

pre-commit.ci autofix

backport or not, could we get this merged please? :)

lief 0.15+ may finally see the large regressions on macos we've been seeing - at least according to the release notes, which have this bullet [bold in original]:

• Fix major performance issue when processing Mach-O binaries on Windows & macOS

Thanks