index : builder | |
Archlinux32 build system | gitolite user |
summaryrefslogtreecommitdiff |
Age | Commit message (Collapse) | Author | |
---|---|---|---|
2023-07-22 | bin/db-update: make step re-runable | Erich Eckner | |
* mkdir needs -p to not complain about already-existing directory | |||
2023-07-17 | bin/sanity-check: execute all left-over intentions at the start | Erich Eckner | |
* the sanity-check should only run, when nothing else runs * this enables some level of clean-up: intentions, that never successfully ran, will be tried again, here | |||
2023-07-17 | bin/sanity-check: do all db operations after getting the lock | Erich Eckner | |
* no need to trouble the db, if we don't run, because we got no lock | |||
2023-07-16 | bin/db-update: use perma_tmp_dir in all intentions, tmp_dir is gone, if the ↵ | Erich Eckner | |
intention needs to be re-executed | |||
2023-07-15 | bin/db-update: fix typo | Erich Eckner | |
2023-07-15 | db-update: make deletion of obsolete pool packages in the db fail the ↵ | Erich Eckner | |
intention, if the db query fails * the `sort -u` was unnecessary anyways * the `sort -u` swallowed the exit code of mysql_run_query * failing mysql_run_query should stop the execution flow, i.e. not remove the intention | |||
2023-07-09 | bin/get-package-updates: less strict locking | Erich Eckner | |
* Do not lock build list, so builders can get jobs. * This may lead to builders getting for jobs, that are currently overwritten or deleted. This should not be worse than if the builder got the job before running get-package-updates, though. | |||
2023-07-08 | schedule-for-rebuild: use correct revisions | Erich Eckner | |
* use null-revisions, if this part is not needed, otherwise make_source_info et al. will fail | |||
2023-07-08 | bin/harvest-commit-times: always order randomly | Erich Eckner | |
2023-07-08 | bin/harvest-commit-times: add force flag to ignore errors | Erich Eckner | |
2023-07-08 | bin/harvest-commit-times: add force flag to ignore errors | Erich Eckner | |
2023-07-08 | bin/harvest-commit-times: add verbosity | Erich Eckner | |
2023-07-01 | renamed pkginfo cache and url, so that build-support works too | Andreas Baumann | |
2023-06-29 | added get-source-info for pkginfo caching | Andreas Baumann | |
2023-06-18 | bin/get-assignment: take max priority of _considered_ build_assignments | Erich Eckner | |
2023-06-18 | rename temp dir to match script name (again) | Erich Eckner | |
2023-06-18 | bin/interpret-mail: call schedule-for-rebuild instead of seed-build-list | Erich Eckner | |
2023-06-18 | bin/return-assignment: call schedule-for-rebuild instead of seed-build-list | Erich Eckner | |
2023-06-18 | bin/schedule-for-rebuild: fix typo | Erich Eckner | |
2023-06-18 | add comment to clarify code | Erich Eckner | |
2023-06-18 | bin/get-package-updates: fix blacklisting | Erich Eckner | |
* the sql wild card for "like" is "%", not "%s" | |||
2023-06-17 | bin/get-assignment: change ordering | Erich Eckner | |
* only consider packages with highest priority * should make "forced-package-builds-*" obsolete | |||
2023-06-17 | bin/get-package-updates: "blacklist" is not a repository, exclude it from ↵ | Erich Eckner | |
the packages32-diff | |||
2023-06-16 | build-support -n <N> should also stop now when do-not-run-build-slave exists | Andreas Baumann | |
2023-06-16 | removed -e in build-packages (-n 0 does the trick already) | Andreas Baumann | |
2023-06-15 | bin/seed-build-list -> bin/schedule-for-rebuild | Erich Eckner | |
* adapt name to what it can do * building completely new packages is not needed anymore * building completely new packages was complex anyways | |||
2023-06-15 | .. | Andreas Baumann | |
2023-06-15 | fixed endless looping in build-packages | Andreas Baumann | |
2023-06-12 | forgot getopt (build-packages) | Andreas Baumann | |
2023-06-12 | build-packages: added a -e parameter for endless self-execution | Andreas Baumann | |
2023-06-11 | extract path to configuration into variable | Erich Eckner | |
* prepare for overwriting location of configuration dir | |||
2023-06-11 | lock on a file in $tmp_dir | Erich Eckner | |
* so we can run the script multiple times in parallel without needing to clone the repository | |||
2023-06-11 | fix comment | Erich Eckner | |
2023-06-11 | make ping-build-master lock on any file given on the command line | Erich Eckner | |
2023-06-11 | removed single quites in build_command (introduced by ${repository} patch) | Andreas Baumann | |
2023-06-09 | bin/generate-key-graph: remove transition keyring | Erich Eckner | |
2023-06-09 | bin/get-package-updates: ignore packages with wrong hash | Erich Eckner | |
2023-06-08 | added ${repository} to build_commands (staging-xxx), for newer devtools32 | Andreas Baumann | |
2023-06-03 | make sanity-check more verbose | Erich Eckner | |
* tell, which architecture is broken | |||
2023-06-03 | make sanity-check more verbose | Erich Eckner | |
* tell, which architecture has inconsistent state | |||
2023-05-24 | fix harvest-commit-times | Erich Eckner | |
make clear, it is a path -> works now for deleted stuff, too | |||
2023-05-24 | shallow clone does not work :-( | Erich Eckner | |
warning: remote HEAD refers to nonexistent ref, unable to checkout fatal: reference is not a tree: 9a167ee2f8b9a0859fbab6cfdc276cf1f272effe ==> ERROR: Failure while creating working copy of gcc git repo Aborting... | |||
2023-05-24 | use bash for PKGBUILD parsing instead of sed-hacking the function names | Erich Eckner | |
2023-05-24 | Revert "revert shallow clone of git sources" | Erich Eckner | |
This reverts commit f026f3947655aa8cb025a9bd04ee0ba567946c03. | |||
2023-05-24 | revert shallow clone of git sources | Erich Eckner | |
* `makepkg --verifysource` does a `git fetch --all -p` * that makes the clone unshallow anyways ... | |||
2023-05-24 | fix shallow git clone of sources | Erich Eckner | |
2023-05-24 | cleanup no-more-working option "--diff|d" from bin/build-packages | Erich Eckner | |
2023-05-23 | try to do a shallow clone before building | Erich Eckner | |
2023-05-23 | fix max verifysource_trial | Erich Eckner | |
2023-05-23 | clean up duplicate logic and unneeded options | Erich Eckner | |