Debian Patches
Status for reprepro/5.4.6+really5.3.2-3
| Patch | Description | Author | Forwarded | Bugs | Origin | Last update |
|---|---|---|---|---|---|---|
| Handle-SIGHUP-like-SIGTERM.patch | Handle SIGHUP like SIGTERM When the terminal reprepro runs on goes away (an ssh session drops, or the process that started reprepro on a pty exits, e.g. a service being restarted or the system shutting down), the kernel sends SIGHUP to reprepro. SIGHUP was not handled, so reprepro died immediately with its Berkeley DB databases open, leaving them corrupt (they are opened without a transactional environment) and leaving a stale db/lockfile. Treat SIGHUP like SIGTERM: finish via the graceful interruption path, which closes the databases and releases the lock. If SIGHUP is inherited as ignored (nohup), leave it ignored so such runs still complete as before. As with the other handled signals, the handler is one-shot: a second SIGHUP still terminates reprepro immediately. |
Jacob Hanks <jacob.hanks@canonical.com> | no | upstream, e4dfe4f2857a3566fa944527156594024009595a | 2026-10-16 | |
| tracking-data-copying.patch | Fix incorrect tracking data when copying packages Tracking database should contain codename, source package name and source version as the key and the containing files as the value in the tracking database. When packages are copied between codenames, it incorrectly uses binary package version, causing inconsistency between the package database and the tracking database. |
Miao Wang <shankerwangmiao@gmail.com> | no | 2026-01-11 |
All known versions for source package 'reprepro'
- 5.5.1-1 (experimental)
- 5.4.6+really5.3.2-3 (sid)
- 5.4.6+really5.3.2-2 (forky)
- 5.4.6+really5.3.2-1+deb13u1 (trixie)
- 5.3.1-1+deb12u2 (bookworm)
