aboutsummaryrefslogtreecommitdiffstatshomepage
path: root/src/server.c (follow)
Commit message (Collapse)AuthorAge
* switch to egor@tensin.nameEgor Tensin2023-12-12
|
* implement a command to list runsEgor Tensin2023-11-15
|
* refactoringEgor Tensin2023-11-12
|
* switch to JSON-RPC as message formatEgor Tensin2023-07-18
| | | | | | Instead of the weird `struct msg` I had, I switched to the JSON-RPC format. It's basically the same, but has a well-defined semantics in case of errors.
* store process output in SQLiteEgor Tensin2023-07-09
|
* test: verify that added runs are in the databaseEgor Tensin2023-07-08
| | | | | And that they're marked as finished. It immediately exposed some concurrency bugs, so some locking has been fixed.
* server: fix a possible leakEgor Tensin2023-07-07
|
* tcp_server: keep track of client threadsEgor Tensin2023-07-05
| | | | | | | This is a major change, obviously; brought to me by Valgrind, which noticed that we don't actually clean up after cimple-client threads. For a more thorough explanation, please see the added comment in tcp_server.c.
* sanitize #include-sEgor Tensin2023-07-04
|
* move custom message parsing to a separate moduleEgor Tensin2023-07-04
|
* storage_sqlite: refactoringEgor Tensin2023-07-04
|
* storage: mark completed runs as suchEgor Tensin2023-07-04
|
* storage: requeue old runs from storage on startupEgor Tensin2023-07-04
|
* tcp_server: always clean up connection descriptorsEgor Tensin2023-07-04
|
* sqlite: store new runs in SQLiteEgor Tensin2023-07-04
|
* storage_sqlite: refactoringEgor Tensin2023-07-04
|
* minor refactoringEgor Tensin2023-06-13
|
* signal: remove the stupid add_to_event_loop wrapperEgor Tensin2023-06-13
|
* server: handle disconnected workers gracefullyEgor Tensin2023-06-13
|
* signal: refactoringEgor Tensin2023-06-13
|
* use signalfd to stop on SIGTERMEgor Tensin2023-06-13
| | | | | | | | | | | | | | | | | | Is this an overkill? I don't know. The thing is, correctly intercepting SIGTERM (also SIGINT, etc.) is incredibly tricky. For example, before this commit, my I/O loops in server.c and worker.c were inherently racy. This was immediately obvious if you tried to run the tests. The tests (especially the Valgrind flavour) would run a worker, wait until it prints a "Waiting for a new command" line, and try to kill it using SIGTERM. The problem is, the global_stop_flag check could have already been executed by the worker, and it would hang forever in recv(). The solution seems to be to use signalfd and select()/poll(). I've never used either before, but it seems to work well enough - at least the very same tests pass and don't hang now.
* msg: rework some APIsEgor Tensin2023-06-11
|
* signal: refactoringEgor Tensin2023-05-15
|
* minor refactoringEgor Tensin2023-05-15
|
* signal: refactoring, add comments in tcp_server, etc.Egor Tensin2023-05-15
|
* EINVAL means EINTR also?Egor Tensin2023-05-15
|
* rework server-worker communicationEgor Tensin2023-05-15
| | | | | | | | | | | | | | | | | | OK, this is a major rework. * tcp_server: connection threads are not detached anymore, the caller has to clean them up. This was done so that the server can clean up the threads cleanly. * run_queue: simple refactoring, run_queue_entry is called just run now. * server: worker threads are now killed when a run is assigned to a worker. * worker: the connection to server is no longer persistent. A worker sends "new-worker", waits for a task, closes the connection, and when it's done, sends the "complete" message and waits for a new task. This is supposed to improve resilience, since the worker-server connections don't have to be maintained while the worker is doing a CI run.
* command: adjust order of parameters to handlersEgor Tensin2023-05-14
|
* msg: add functions for one-off communicationEgor Tensin2023-05-14
|
* ci_queue -> run_queueEgor Tensin2023-05-13
| | | | Also, some minor refactoring.
* command: refactoringEgor Tensin2023-05-13
|
* best practices & coding style fixesEgor Tensin2023-05-13
| | | | | | | | * I don't really need to declare all variables at the top of the function anymore. * Default-initialize variables more. * Don't set the output parameter until the object is completely constructed.
* add command module to handle request-response communicationsEgor Tensin2023-05-13
|
* ci_queue: rename a couple of functionsEgor Tensin2023-05-12
|
* shut down server/workers gracefully on SIGTERMEgor Tensin2023-05-06
|
* get rid of __attribute__((constructor))Egor Tensin2023-05-06
| | | | Explicit is better than implicit.
* make struct ci_queue_entry opaqueEgor Tensin2023-04-29
|
* make struct server opaqueEgor Tensin2023-04-29
|
* make struct tcp_server opaqueEgor Tensin2023-04-29
|
* fix a typoEgor Tensin2023-04-27
|
* rename commandsEgor Tensin2023-04-27
|
* add copyright noticesEgor Tensin2022-12-02
|
* create SQLite database on startupEgor Tensin2022-09-11
|
* log: refactoringEgor Tensin2022-09-08
|
* sanitize #include-sEgor Tensin2022-09-08
|
* update command namesEgor Tensin2022-08-28
|
* server: notify workers about requeued jobsEgor Tensin2022-08-28
| | | | This allows free workers to pick up jobs after dead workers.
* server: notify all threads about shutting downEgor Tensin2022-08-28
| | | | | | | The problem is pthread_cond_destroy is unsafe to call if there're threads waiting in pthread_cond_wait. I'm not sure this fix is enough: what if the "broadcast" doesn't reach the threads until we call pthread_cond_destroy? Does it even work that way? Idk
* make proper "error" messagesEgor Tensin2022-08-28
| | | | | Previously, the client had no way to distinguish errors from succesful calls.
* make compilers happierEgor Tensin2022-08-28
|
generated by cgit v1.2.3 (git 2.25.1) at 2024-09-19 16:49:36 +0000