Commit message (Collapse) | Author | ||
---|---|---|---|
2023-07-19 | move valgrind.sh to scripts/ | Egor Tensin | |
2023-07-18 | net: don't copy data in struct buf | Egor Tensin | |
2023-07-18 | process: preserve the exit code more | Egor Tensin | |
If the process is killed by a signal, preserve the exit code as it would be reported by $?. | |||
2023-07-18 | json_rpc: removed unused routines | Egor Tensin | |
2023-07-18 | net: remove unused buf_* routines | Egor Tensin | |
2023-07-18 | switch to JSON-RPC as message format | Egor Tensin | |
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. | |||
2023-07-11 | test: add test for segfaulting CI script | Egor Tensin | |
The C code leaked out of src/, so I moved .clang-format and some compile options to the root directory. Also, I'm starting to hit test execution limits; I'm going to limit the repositories used for stress testing. | |||
2023-07-11 | sqlite: make some identifiers more readable | Egor Tensin | |
2023-07-11 | sqlite: rename SQL schema variables | Egor Tensin | |
2023-07-10 | string: add a comment | Egor Tensin | |
2023-07-10 | test: test long CI run output | Egor Tensin | |
It immediately exposed a horrible bug in net.c, which is now fixed. | |||
2023-07-09 | store process output in SQLite | Egor Tensin | |
2023-07-08 | sqlite: SQL formatting | Egor Tensin | |
2023-07-08 | test: verify that added runs are in the database | Egor Tensin | |
And that they're marked as finished. It immediately exposed some concurrency bugs, so some locking has been fixed. | |||
2023-07-07 | clang-format: don't break string literals | Egor Tensin | |
2023-07-07 | test: add some basic command-line usage tests | Egor Tensin | |
2023-07-07 | support code coverage report generation | Egor Tensin | |
2023-07-07 | server: fix a possible leak | Egor Tensin | |
2023-07-07 | sqlite: minor refactoring | Egor Tensin | |
2023-07-06 | storage_sqlite: avoid races when inserting repos | Egor Tensin | |
Insert first, then query the ID. That way, it'll work even if there're other workers inserting new repos. | |||
2023-07-06 | cmake: fix Valgrind runs with Clang builds | Egor Tensin | |
2023-07-05 | worker: actually stay offline | Egor Tensin | |
I don't know what I was thinking, but contrary to my intention, the worker stayed connected to the server all the time. | |||
2023-07-05 | tcp_server: minor refactoring | Egor Tensin | |
2023-07-05 | tcp_server: close client connections | Egor Tensin | |
I kinda forgot that I'm supposed to clean up client connections, and they immediately blew up in my face. | |||
2023-07-05 | command: dedupe the code | Egor Tensin | |
2023-07-05 | tcp_server: keep track of client threads | Egor Tensin | |
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. | |||
2023-07-04 | sanitize #include-s | Egor Tensin | |
2023-07-04 | worker: close the leftover descriptor | Egor Tensin | |
Thanks, Valgrind! As a note: if I think that Valgrind reports a false positive, chances are, it's not. | |||
2023-07-04 | move custom message parsing to a separate module | Egor Tensin | |
2023-07-04 | storage_sqlite: refactoring | Egor Tensin | |
2023-07-04 | sqlite: rename a column | Egor Tensin | |
2023-07-04 | storage: mark completed runs as such | Egor Tensin | |
2023-07-04 | storage: requeue old runs from storage on startup | Egor Tensin | |
2023-07-04 | tcp_server: always clean up connection descriptors | Egor Tensin | |
2023-07-04 | sqlite: store new runs in SQLite | Egor Tensin | |
2023-07-04 | storage_sqlite: refactoring | Egor Tensin | |
2023-07-04 | sqlite: add run status field | Egor Tensin | |
2023-07-04 | fix a compiler warning | Egor Tensin | |
2023-07-04 | appease clang-format | Egor Tensin | |
2023-06-30 | use designated struct initializers more | Egor Tensin | |
2023-06-30 | show git hash with --version | Egor Tensin | |
Also, use cmake's configure_file to build string constants in. | |||
2023-06-30 | fix function names in error messages | Egor Tensin | |
2023-06-30 | use __WNOTHREAD w/ waitpid | Egor Tensin | |
I learned about this flag on my previous job; basically, it should be enabled by default IMO, so that the thread doesn't receive signals from other threads' children. Here, it doesn't matter too much, since we're waiting for a specific child. However, if we were to use waitpid(-1, ...), it would be essential to use this flag. Still, even here, it's good to have this on. | |||
2023-06-28 | log: minor refactoring | Egor Tensin | |
2023-06-28 | process: fix function name in logs | Egor Tensin | |
2023-06-28 | ci: environment array to execvpe must be NULL-terminated | Egor Tensin | |
Thanks, Valgrind. | |||
2023-06-28 | log: support logging levels | Egor Tensin | |
2023-06-28 | test: skip ci.sh w/ Valgrind | Egor Tensin | |
2023-06-28 | valgrind: exit with an error after leak detection | Egor Tensin | |
I cannot believe this isn't default. | |||
2023-06-28 | file: readlink(2) doesn't null-terminate | Egor Tensin | |
Thanks, valgrind. |