Debian Patches
Status for procserv/2.8.0-2
| Patch | Description | Author | Forwarded | Bugs | Origin | Last update |
|---|---|---|---|---|---|---|
| 0001-Use-long-int-for-coresize-argument.patch | Use long int for --coresize argument Since the variable holding the core size is an `int`, users of --coresize will have core dumps truncated and hence unusable if it is greater than the size can hold (~2GB for most 64-bits platforms). Use `long` to increase it to 64-bits on platforms that follow LP64 (most Unix systems), providing enough room for any reasonable core size, while keeping it limited to 2GB on LLP64 (such as MS Windows) [1]. [1] https://unix.org/whitepapers/64bit.html |
Ralph Lange <ralph.lange@gmx.de> | yes | upstream | upstream, https://github.com/ralphlange/procServ/commit/498687f49adf008ea206969ce06b4ce99c453d0d | 2020-07-01 |
| 0002-Report-error-if-setsockopt-fails.patch | Report error if setsockopt() fails | Martin Konrad <konrad@frib.msu.edu> | no | upstream, https://github.com/ralphlange/procServ/commit/9078191e8c19d5063a4da738ba8b46eab9f26c78 | 2020-07-08 | |
| 0003-Modify-restartMode-oneshot-feature-to-run-once-AFTER.patch | Modify restartMode oneshot feature to run once AFTER selecting oneshot restartMode. If IOC process is not running and the developer tries to change restartMode, it kills procServ when the developer toggles through oneshot mode. This commit sets firstRun true as restartMode is toggled. Thus, developers can toggle through oneshot mode w/o killing procServ and the IOC will run at least once before oneshot mode will kill procServ. |
Bruce Hill <bhill@slac.stanford.edu> | no | upstream, https://github.com/ralphlange/procServ/commit/66137916c0d88dabc21d923f10468bcb1503e7ce | 2020-09-08 | |
| 0004-Write-pidfile-of-child-in-parent.patch | Write pidfile of child in parent Avoids race condition if using a Type=forking systemd service |
=?utf-8?q?Kriszti=C3=A1n_L=C3=B6ki?= <krisztian.loki@ess.eu> | no | upstream, https://github.com/ralphlange/procServ/commit/9e91bd0db0b292d45947d2cce4fdc25a7fcafc35 | 2021-03-04 | |
| 0005-Close-logfile-FD-for-child-processes.patch | Close logfile FD for child processes When the child processes are spawned, they inherit the logfile file descriptor. If the procserv logs are rotated using the logrotate utility, then the child processes still have a open file handle on the original logfile before it gets rotated. This prevents the kernel from freeing up the space used even though the original logfile has been deleted. Over time, or with large logfiles, this can result in a lot of phantom space being used up in the log directory. This PR closes the logfile file descriptor before exec()-ing the child processes, allowing the logfile to be rotated and cleaned up by the kernel. |
Michael Thomas <mthomas@ligo-la.caltech.edu> | no | upstream, https://github.com/ralphlange/procServ/commit/fc3b770611ed4755f3f09986f64519e75e3bad4a | 2023-11-08 | |
| 0006-Clarify-licensing-terms-in-documentation-and-README.patch | Clarify licensing terms in documentation and README Triggered by a discussion on the Debian packaging, see https://salsa.debian.org/debian/procserv/-/merge_requests/4 |
Ralph Lange <ralph.lange@gmx.de> | no | upstream, https://github.com/ralphlange/procServ/commit/52fc9bc4a5c437a8fd3fbb56c21e5866f906815c | 2024-11-04 | |
| 0007-Don-t-install-manpage-source-with-disabled-docs.patch | Don't install manpage source with disabled docs Users should be able to disable documentation installation with `configure --enable-doc=no`. However, procServ.txt is still installed when that flag is provided. Teach automake how to properly honor the --enable-doc=no option by moving the manpage source file along with its generated files, guarded by `if INSTALL_DOC`. |
"Henrique F. Simoes" <henriquesimoes@riseup.net> | no | 2025-02-09 |