bernardo/Nomarchy
1
1
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

fix(idle): don't lock before an encrypted hibernate (was: unlock after)

Browse Source 
The previous nomarchy-hibernate-unlock ran `pkill hyprlock` after a
hibernate resume, but killing a Wayland session-lock client without
releasing the lock trips hyprlock's "go to a tty" crash failsafe -- the
compositor keeps the screen locked for safety. So resume showed a hyprlock
error screen instead of unlocking.

Fix it the right way: never engage the lock before an encrypted hibernate
(the LUKS passphrase at resume is the gate). Replace the post-resume unlock
with a nomarchy-lock-before-sleep unit that takes over hypridle's
before_sleep_cmd: it locks on the RAM-resume sleeps (suspend / hybrid-sleep
/ suspend-then-hibernate) always, and on hibernate.target only when the disk
is unencrypted. idle.nix drops before_sleep_cmd accordingly.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
This commit is contained in:
Bernardo Magri
2026-06-15 10:15:31 +01:00
parent a606c0d498
commit 1b0eeeaf6c
3 changed files with 40 additions and 27 deletions
Download Patch File Download Diff File Expand all files Collapse all files

39
modules/nixos/default.nix
View File

@@ -123,23 +123,32 @@ in
networking.networkmanager.enable = lib.mkDefault true;
# No double-unlock on hibernate. hypridle locks hyprlock before every
# sleep (right for suspend), but a hibernate resume is already gated by
# the LUKS passphrase entered at boot — so the user would type a password
# twice. Once we're back from an *encrypted* hibernate, drop the now-
# redundant hyprlock. Ordered After the hibernate service, so it runs
# post-resume; pulled in only by hibernate.target (suspend, and the
# RAM-resume phase of suspend-then-hibernate, stay locked — they have no
# passphrase gate). Gated on LUKS: an unencrypted hibernate keeps its lock.
systemd.services.nomarchy-hibernate-unlock =
lib.mkIf (builtins.attrNames config.boot.initrd.luks.devices != [ ]) {
description = "Dismiss hyprlock after resuming from an encrypted hibernate";
after = [ "systemd-hibernate.service" ];
wantedBy = [ "hibernate.target" ];
# No double-unlock on hibernate. Locking the session before sleep is
# right for suspend (resumes from RAM, no other gate), but an encrypted
# hibernate already resumes through the LUKS passphrase — a hyprlock on
# top is a second password. And we can't just drop the lock after the
# resume: a Wayland session-lock whose client dies without releasing
# trips hyprlock's "go to a tty" crash failsafe (the compositor keeps the
# screen locked for safety). So don't lock before an encrypted hibernate
# in the first place. hypridle can't tell suspend from hibernate, hence a
# system unit hooked to the sleep targets: it locks on the RAM-resume
# sleeps always, and on hibernate only when the disk is unencrypted (no
# LUKS gate to rely on). Replaces hypridle's old before_sleep_cmd.
systemd.services.nomarchy-lock-before-sleep =
let
encrypted = builtins.attrNames config.boot.initrd.luks.devices != [ ];
# Sleeps whose normal resume is from RAM — always lock these.
ramTargets = [ "suspend.target" "hybrid-sleep.target" "suspend-then-hibernate.target" ];
ramServices = [ "systemd-suspend.service" "systemd-hybrid-sleep.service" "systemd-suspend-then-hibernate.service" ];
in
{
description = "Lock the session before sleep (skipped for an encrypted hibernate)";
before = ramServices ++ lib.optional (!encrypted) "systemd-hibernate.service";
wantedBy = ramTargets ++ lib.optional (!encrypted) "hibernate.target";
serviceConfig = {
Type = "oneshot";
# `-`: a missing hyprlock (idle disabled, or not locked) isn't a failure.
ExecStart = "-${pkgs.procps}/bin/pkill -x hyprlock";
# Plural: a system unit has no session of its own to lock.
ExecStart = "${config.systemd.package}/bin/loginctl lock-sessions";
};
};