make-symlinks-relative: fix no such file or directory if output is cr…

[Artturin]

…eated in postFixup

wlroots(and others) have

wlroots> +++ find /nix/store/3a0xwszw8n5dzzhsgfnilvsqi4hk565s-wlroots-0.15.1-examples -type l -print0
wlroots> find: '/nix/store/3a0xwszw8n5dzzhsgfnilvsqi4hk565s-wlroots-0.15.1-examples': No such file or directory

because the examples output is created in postFixup while this hook runs in fixupPhase

Description of changes

Things done

• Built on platform(s)
  • x86_64-linux
  • aarch64-linux
  • x86_64-darwin
  • aarch64-darwin
• For non-Linux: Is sandbox = true set in nix.conf? (See Nix manual)
• Tested, as applicable:
  • NixOS test(s) (look inside nixos/tests)
  • and/or package tests
  • or, for functions and "core" functionality, tests in lib/tests or pkgs/test
  • made sure NixOS tests are linked to the relevant packages
• Tested compilation of all packages that depend on this change using nix-shell -p nixpkgs-review --run "nixpkgs-review rev HEAD". Note: all changes have to be committed, also see nixpkgs-review usage
• Tested basic functionality of all binary files (usually in ./result/bin/)
• 23.05 Release Notes (or backporting 22.11 Release notes)
  • (Package updates) Added a release notes entry if the change is major or breaking
  • (Module updates) Added a release notes entry if the change is significant
  • (Module addition) Added a release notes entry if adding a new NixOS module
  • (Release notes changes) Ran nixos/doc/manual/md-to-db.sh to update generated release notes
• Fits CONTRIBUTING.md.

[andir]

Maybe those builds are wrong then? Isn't it rather postBuild or postInstall or installPhase where outputs should be written? Doing that in fixup sounds wrong.

[Artturin]

Maybe those builds are wrong then? Isn't it rather postBuild or postInstall or installPhase where outputs should be written? Doing that in fixup sounds wrong.

Yeah maybe but wouldn't it be better for this hook to run on symlinks added in postFixup

Afaik there have been no issues caused by this hook

[Artturin]

@andir anymore questions or do you think this is ready to merge?

[Artturin]

@ofborg build sway zsh

[andir]

@andir anymore questions or do you think this is ready to merge?

I haven't tested this, so I can't really give a +1 just yet. I'll try to test this on the weekend.

[Artturin]

i added a test for the hook

$ nix build -f . "tests.hooks"
this derivation will be built:
  /nix/store/askavvq9j724816kv9j5dkncp7c4z0p3-test-make-symlinks-relative.drv
building '/nix/store/askavvq9j724816kv9j5dkncp7c4z0p3-test-make-symlinks-relative.drv'...
test-make-symlinks-relative> symlink before patching: /nix/store/agabiwibxpvkfl4aa3ilp7v1mimxi744-test-make-symlinks-relative/bar/foo
test-make-symlinks-relative> rewriting symlink /nix/store/agabiwibxpvkfl4aa3ilp7v1mimxi744-test-make-symlinks-relative/baz/foo to be relative to /nix/store/agabiwibxpvkfl4aa3ilp7v1mimxi744-test-make-symlinks-relative
test-make-symlinks-relative> symlink after patching: ../bar/foo
test-make-symlinks-relative> symlink isn't broken
test-make-symlinks-relative> absolute symlink was made relative

[Artturin]

also tested with

nix build ".#bzip2^*"
...
bzip2> rewriting symlink /nix/store/8dvdlr36j65b684ba1r160pdya6i329f-bzip2-1.0.8/lib/libbz2.so.1.0 to be relative to /nix/store/8dvdlr36j65b684ba1r160pdya6i329f-bzip2-1.0.8

$ \ls -l ./result/lib
total 43
-r-xr-xr-x  2 root root   946  1. 1.  1970 libbz2.la
lrwxrwxrwx 67 root root    15  1. 1.  1970 libbz2.so -> libbz2.so.1.0.8
lrwxrwxrwx 67 root root    15  1. 1.  1970 libbz2.so.1 -> libbz2.so.1.0.8
lrwxrwxrwx 67 root root    15  1. 1.  1970 libbz2.so.1.0 -> libbz2.so.1.0.8
-r-xr-xr-x  2 root root 79072  1. 1.  1970 libbz2.so.1.0.8

with dontRewriteSymlinks = true;

$ \ls -l ./result/lib
total 51
-r-xr-xr-x  2 root root   946  1. 1.  1970 libbz2.la
lrwxrwxrwx 69 root root    15  1. 1.  1970 libbz2.so -> libbz2.so.1.0.8
lrwxrwxrwx 69 root root    15  1. 1.  1970 libbz2.so.1 -> libbz2.so.1.0.8
lrwxrwxrwx  2 root root    75  1. 1.  1970 libbz2.so.1.0 -> /nix/store/fd3jygv5iflmdflkwr9g0gwg79v7r65q-bzip2-1.0.8/lib/libbz2.so.1.0.8
-r-xr-xr-x  2 root root 78232  1. 1.  1970 libbz2.so.1.0.8

[SuperSandro2000]

Maybe those builds are wrong then? Isn't it rather postBuild or postInstall or installPhase where outputs should be written? Doing that in fixup sounds wrong.

Then we should at least error out in such scenarios instead of silently continuing.

[andir]

Maybe those builds are wrong then? Isn't it rather postBuild or postInstall or installPhase where outputs should be written? Doing that in fixup sounds wrong.

Then we should at least error out in such scenarios instead of silently continuing.

Feel free. I'll yield on this issue. You seem to know what you are doing.

Don't mention me in this PR again. Thanks.

[Artturin]

Feel free. I'll yield on this issue. You seem to know what you are doing.

Don't mention me in this PR again. Thanks.

Alright, thanks for the hook.

[Artturin]

@ofborg build tests.hooks

[Artturin]

Mention me if there are issues