coreutils: 9.3 -> 9.4

[trofi]

Changes: https://lists.gnu.org/archive/html/info-gnu/2023-08/msg00007.html

Description of changes

Things done

• Built on platform(s)
  • x86_64-linux
  • aarch64-linux
  • x86_64-darwin
  • aarch64-darwin
• For non-Linux: Is sandboxing enabled in nix.conf? (See Nix manual)
  • sandbox = relaxed
  • sandbox = true
• 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/)
• 23.11 Release Notes (or backporting 23.05 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
• Fits CONTRIBUTING.md.

[dasJ]

See also #252165 which also drops some patches. Can you drop them here as well?

[trofi]

See also #252165 which also drops some patches. Can you drop them here as well?

AFAIU those were dropped around 9.0 -> 9.1 update (and a bit pater). coreutils does not apply any patches in 9.3. I dropped aarch64-specific sed '2i print "Skipping tail assert test"; exit 77' -i ./tests/tail-2/assert.sh from there.

[Artturin]

Fedora has added only one new patch for 9.4 which is related to --enable-systemd which we don't do so not relevant for us
https://src.fedoraproject.org/rpms/coreutils/blob/rawhide/f/coreutils-9.4-systemd-coredump.patch

[minijackson]

I am getting some errors with coreutils 9.4 when using stdenv.mkDerivation with src = ./some/local/dir, during unpackPhase.

# With 9.3 (tried it manually outside the Nix sandbox)
/nix/store/59mv7mybz16jjvnkz4z5shqgaqm4k009-coreutils-full-9.3/bin/cp -pr --reflink=auto -- /nix/store/...-vim vim
# Success!

# With 9.4 (tried both inside and outside the Nix sandbox)
/nix/store/ifzwv2xqwdnv1gz87rxkizi67py5p3vj-coreutils-9.4/bin/cp -pr --reflink=auto -- /nix/store/...-vim vim
# Error:
# /nix/store/ifzwv2xqwdnv1gz87rxkizi67py5p3vj-coreutils-9.4/bin/cp: setting permissions for 'vim': Invalid argument

Can anyone confirm? I tried it on an ext4 filesystem.

[minijackson]

After a running both under strace, I can see this from 9.3:

mkdirat(AT_FDCWD, "vim", 0555)          = 0
newfstatat(AT_FDCWD, "vim", {st_mode=S_IFDIR|0555, st_size=4096, ...}, AT_SYMLINK_NOFOLLOW) = 0
openat(AT_FDCWD, "vim", O_RDONLY|O_NOFOLLOW|O_CLOEXEC|O_PATH) = 3
newfstatat(3, "", {st_mode=S_IFDIR|0555, st_size=4096, ...}, AT_EMPTY_PATH) = 0
chmod("/proc/self/fd/3", 040755)        = 0
close(3)                                = 0

And this from 9.4:

mkdirat(AT_FDCWD, "vim", 0555)          = 0
newfstatat(AT_FDCWD, "vim", {st_mode=S_IFDIR|0555, st_size=4096, ...}, AT_SYMLINK_NOFOLLOW) = 0
fchmodat2(AT_FDCWD, "vim", 040755, AT_SYMLINK_NOFOLLOW) = -1 EINVAL (Invalid argument)

It seems they changed the way they chmod directories. According to the fchmodat manpage, EINVAL is returned when "Invalid flag specified in flags", which here is AT_SYMLINK_NOFOLLOW.

Above we can see:

AT_SYMLINK_NOFOLLOW: If pathname is a symbolic link, do not dereference it: instead operate on the link itself. This flag is not currently implemented.

This makes me confused…

[minijackson]

Made a separate issue here #302046