UI deadlock after a missed remote backup fails with a pre-backup command

[signalwerk]

Hello, hello!
Thank you for Vorta! It's great. But I recently bought a NAS, and since then I've noticed that Vorta freezes, and I didn't realize what the pattern was. I pointed an LLM at it to analyze what's the problem. And I can confirm it found the smoking gun. Here we go;

--written by the LLM from here on--

Description:

Environment: Vorta 0.11.5, Borg 1.4.4, macOS 15.7.4.

Reproduction:

1. Configure a scheduled SSH/NAS backup.
2. Enable “Run missed backups on startup or wakeup.”
3. Add a pre-backup command, for example /bin/sleep 5.
4. Complete one successful backup, then quit Vorta until an interval is missed.
5. Connect to a network with Internet access where the NAS hostname is unreachable.
6. Start Vorta.

Vorta runs the missed backup, Borg fails, and the scheduler immediately retries. During the next pre-backup command, Vorta’s main thread deadlocks. The UI becomes unresponsive. Repository-lock dialogs also freeze, so neither “Cancel” nor “Break the lock” works.

A process sample showed the main thread inside QCoreApplication.processEvents(), blocked on lock_PyThread_acquire_lock.

Expected: the backup fails or is postponed while Vorta remains responsive.

[m3nu]

This looks like the same lock re-entry as #2367; the fix landed on master in #2558 (fcccbbb). It will be in the next release. If you can run from master in the meantime, please report whether the freeze on the retried pre-backup command is gone.