treewide: stdenv.is -> stdenv.hostPlatform.is

[JohnRTitor]

Commands ran:

fd --type f "\.nix" | xargs sd --fixed-strings "stdenv.is" "stdenv.hostPlatform.is"
fd --type f "\.nix" | xargs sd --fixed-strings "stdenv'.is" "stdenv'.hostPlatform.is"
fd --type f "\.nix" | xargs sd --fixed-strings "clangStdenv.is" "clangStdenv.hostPlatform.is"
fd --type f "\.nix" | xargs sd --fixed-strings "gccStdenv.is" "gccStdenv.hostPlatform.is"
fd --type f "\.nix" | xargs sd --fixed-strings "stdenvNoCC.is" "stdenvNoCC.hostPlatform.is"
fd --type f "\.nix" | xargs sd --fixed-strings "inherit (stdenv) is" "inherit (stdenv.hostPlatform) is"
fd --type f "\.nix" | xargs sd  "buildStdenv.is" "buildStdenv.hostPlatform.is"
fd --type f "\.nix" | xargs sd  "effectiveStdenv.is" "effectiveStdenv.hostPlatform.is"
fd --type f "\.nix" | xargs sd  "originalStdenv.is" "originalStdenv.hostPlatform.is"

Following #341407

stdenv.isLinux -> stdenv.hostPlatform.isLinux, is pretty much easy expansion of alias. This should cause no rebuilds.

However some packages have stdenv.isLinux in doCheck, meaning package tests have to be run when building in linux. So instead of hostPlatform there use buildPlatform there. This could cause some rebuilds.

---

Add a 👍 reaction to pull requests you find important.

[Artturin]

These sneaked back in 5136e37 lol

[Artturin]

Just do the commands in #341407 instead of only isLinux

[emilazy]

I am unsure about the second commit here. Is it going to work with cross? Don’t we want a canExecute thing instead?

[Artturin]

I am unsure about the second commit here. Is it going to work with cross? Don’t we want a canExecute thing instead?

nixpkgs/pkgs/stdenv/generic/make-derivation.nix

Line 249 in cac8a72

doCheck' = doCheck && stdenv.buildPlatform.canExecute stdenv.hostPlatform;

canExecute is added by stdenv on top

However not sure if it's added when doing overrideAttrs like done in lua-modules

[JohnRTitor]

I am unsure about the second commit here. Is it going to work with cross?

It should work I suppose, the checks needs to be enabled on the build machine, not on the host machine where it would run.

OKAY I fundamentally misunderstood I believe, let's change it back to hostPlatform

[emilazy]

Formatter is unhappy.

[JohnRTitor]

Not related to this PR, the errors are occuring as the original files are not formatted. We are only touching a line, formatting them here is out of scope

[emilazy]

CI only complains if files that were formatted before the PR become unformatted, so I don’t think that’s true.

[jopejoe1]

It complains about pkgs/applications/radio/sdrangel/default.nix, pkgs/by-name/er/erlang-language-platform/package.nix and pkgs/by-name/im/imhex/package.nix not being formated.

[JohnRTitor]

Yeah, it seems like nixfmt will throw errors if line char limits have been exceeded and use an alternate style.

Done, this should make the CI happy, I have still decided to make the commit seperately though, for easier skimming through the diff. The PR can be squash merged if needed.

[JohnRTitor]

Eval successfully passed but a merge conflict is shown. Intend to merge after rebasing, as while we race for the eval to complete again likely another conflict will arise