Hmm. I'll give this a try and try either this or revert, whichever is better.

Success on aarch64-linux (full log)

/nix/store/z7k4qnrvab9l08aynamhwq27m36g23hz-strace-4.21

Failure on x86_64-darwin (full log)

Package ‘strace-4.21’ in /private/var/lib/ofborg/checkout/repo/38dca4e3aa6bca43ea96d2fcc04e8229/builder/lnl7-mac/pkgs/development/tools/misc/strace/default.nix:20 is not supported on ‘x86_64-darwin’, refusing to evaluate.

a) For `nixos-rebuild` you can set
  { nixpkgs.config.allowBroken = true; }
in configuration.nix to override this.

b) For `nix-env`, `nix-build`, `nix-shell` or any other Nix command you can add
  { allowBroken = true; }
to ~/.config/nixpkgs/config.nix.

Success on x86_64-linux (full log)

/nix/store/lscmjgcp9g7s44wras5vjn3bp76vp768-strace-4.21

Hmm. I'll give this a try and try either this or revert, whichever is better.

Not sure I follow, but are you saying you'll take things from here? :)

Not sure I follow, but are you saying you'll take things from here? :)

Yep. :)

I googled around a bit and basically I think it's an upstream bug that the build fails fatally (by default) when it finds out that -m32 doesn't work on AArch64, given that GCC doesn't support such a flag there (and I expect they never will). So I hope this can be removed one day in a further upstream release.

I did do s/=no/=check/
to make it a bit more futureproof (i.e. in case strace grows support for -milp32 on AArch64 one day, we don't want to unconditionally disable it): 8989a6b

Perfect, I wasn't thrilled about the "=no" bit myself, check (assuming it does what it sounds, how is this not the default??) sounds great.

Thanks!