staging-next 2023-05-02

[vcunat]

According to 23.05 schedule, this this iteration is the last one allowed to contain breaking changes (except desktop environments).

Helpful links

• workflow docs: https://nixos.org/manual/nixpkgs/unstable/#submitting-changes-commit-policy
• important jobs:
  https://hydra.nixos.org/job/nixpkgs/staging-next/unstable#tabs-constituents
  https://hydra.nixos.org/job/nixos/staging-next-small/tested
• jobsets:
  https://hydra.nixos.org/jobset/nixpkgs/staging-next
  https://hydra.nixos.org/jobset/nixos/staging-next-small
• nix-review reports: https://malob.github.io/nix-review-tools-reports/
• previous staging-next: staging-next 2023-04-27 #228476

---

Mass breakages

(will be edited based on progress)

• python311Packages.pytest-benchmark (all platforms but 3.11 only)
• notmuch
  https://hydra.nixos.org/build/218564125#tabs-buildsteps
  1d6af3f

[vcunat]

Perhaps notmuch maintainers could look into its build regression? @flokli @puckipedia

EDIT: logs differ by platform, apparently:

• x86_64-linux https://hydra.nixos.org/log/67n5cpzsij65gqf8sv6yfz1q023mv7sb-notmuch-0.37.drv
• aarch64-linux https://hydra.nixos.org/log/6qa6a5i49xvzddyppx4nsw4wxyv5x5bg-notmuch-0.37.drv

[vcunat]

So the gnupg patch-update (in 122e774) certainly does break notmuch build (on aarch64-linux). No idea why.

There were at least two different notmuch breakages in the bisected span (even though it's just over several days), so it might be more complicated... though I think this commit introduced the right error (same-looking log as in latest staging-next).

EDIT: I did it on aarch64.nixos.community so it has the binaries, in case someone wants to play with this.

[vcunat]

notmuch: relevant parts of test logs (thanks to Lily):

T350-crypto: Testing PGP/MIME signature verification and decryption
 PASS   emacs delivery of signed message via fcc
 PASS   emacs delivery of signed message via fcc and smtp
 PASS   signed part content-type indexing
 PASS   signature verification
 PASS   detection of modified signed contents
 PASS   corrupted pgp/mime signature
 PASS   signature verification without full user ID validity
 PASS   signature verification with signer key unavailable
FATAL: /build/notmuch-0.37/test/T350-crypto.sh: interrupted by signal 15

[...]
T357-index-decryption: Testing indexing decrypted mail
FATAL: /build/notmuch-0.37/test/T357-index-decryption.sh: interrupted by signal 15

[vcunat]

Well, the bad tests are skipped in notmuch now, but I think it would be still nice if someone looked into that...

[vcunat]

• sway test times out when waiting for gpg to show "Passphrase"

    # Test gpg-agent starting pinentry-gnome3 via D-Bus (tests if
    # $WAYLAND_DISPLAY is correctly imported into the D-Bus user env):
    swaymsg("exec gpg --no-tty --yes --quick-generate-key test")
    machine.wait_until_succeeds("pgrep --exact gpg")
    machine.wait_for_text("Passphrase")
    machine.screenshot("gpg_pinentry")
    machine.send_key("alt-shift-q")
    machine.wait_until_fails("pgrep --exact gpg")

  This is blocking nixos-unstable channel.