aboutsummaryrefslogtreecommitdiffstatshomepage
Commit message (Collapse)AuthorAge
* test: refactor command line optionsEgor Tensin2023-07-19
|
* README: updateEgor Tensin2023-07-19
|
* workflows/ci: restructure job dependenciesEgor Tensin2023-07-19
| | | | | The primary reason behind this was the desire to get rid of duplicate deployments to GitHub Pages.
* README: updateEgor Tensin2023-07-18
|
* workflows/ci: publish stuff to gh-pagesEgor Tensin2023-07-18
|
* test: store multiple flame graphs alongsideEgor Tensin2023-07-18
| | | | | | | Previously, it would get stored in build/flame_graph/flame_graphs.svg. Now, the test repository codename is added to the file name. Also, some refactoring and simplifying test filtering.
* flame_graph.sh -> flamegraph.shEgor Tensin2023-07-18
| | | | Inspired by flamegraph.pl.
* test: minor refactoringEgor Tensin2023-07-18
|
* test: exclude silly tests from python_tests_valgrindEgor Tensin2023-07-18
|
* test: speed up output_long test_repo testsEgor Tensin2023-07-18
|
* implement flame graph generationEgor Tensin2023-07-18
|
* net: don't copy data in struct bufEgor Tensin2023-07-18
|
* process: preserve the exit code moreEgor Tensin2023-07-18
| | | | | If the process is killed by a signal, preserve the exit code as it would be reported by $?.
* json_rpc: removed unused routinesEgor Tensin2023-07-18
|
* net: remove unused buf_* routinesEgor Tensin2023-07-18
|
* 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.
* docker: sanitize package dependenciesEgor Tensin2023-07-14
|
* test: move some code to lib/Egor Tensin2023-07-12
|
* test: use namedtuple where appropriateEgor Tensin2023-07-12
|
* test: add test for segfaulting CI scriptEgor Tensin2023-07-11
| | | | | | | | 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.
* sqlite: make some identifiers more readableEgor Tensin2023-07-11
|
* sqlite: rename SQL schema variablesEgor Tensin2023-07-11
|
* test: more compact CI scriptsEgor Tensin2023-07-10
|
* test: fix Python freezesEgor Tensin2023-07-10
| | | | | I would get random freezes when running tests; I completely forgot that logging & multiprocessing don't play well together.
* test: code styleEgor Tensin2023-07-10
|
* test: test CI run output containing \0Egor Tensin2023-07-10
|
* string: add a commentEgor Tensin2023-07-10
|
* test: test long CI run outputEgor Tensin2023-07-10
| | | | It immediately exposed a horrible bug in net.c, which is now fixed.
* test: prettier test namesEgor Tensin2023-07-09
|
* test: test empty CI run output, refactoringEgor Tensin2023-07-09
| | | | | Meh, moving the shell code generation to Python I like better in the end.
* store process output in SQLiteEgor Tensin2023-07-09
|
* test: attempt to fix random port selection againEgor Tensin2023-07-09
|
* test: print test timingsEgor Tensin2023-07-09
|
* test: increase timeoutsEgor Tensin2023-07-09
| | | | All the new added tests require more time.
* test: don't wait for hanging processesEgor Tensin2023-07-09
|
* test: fix sqlite3.connect usageEgor Tensin2023-07-09
|
* sqlite: SQL formattingEgor Tensin2023-07-08
|
* test: dedupe code by using @parametrizeEgor Tensin2023-07-08
|
* test: refactoringEgor Tensin2023-07-08
|
* 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.
* clang-format: don't break string literalsEgor Tensin2023-07-07
|
* test: add some basic command-line usage testsEgor Tensin2023-07-07
|
* support code coverage report generationEgor Tensin2023-07-07
|
* Makefile: add release targetEgor Tensin2023-07-07
|
* server: fix a possible leakEgor Tensin2023-07-07
|
* sqlite: minor refactoringEgor Tensin2023-07-07
|
* storage_sqlite: avoid races when inserting reposEgor Tensin2023-07-06
| | | | | Insert first, then query the ID. That way, it'll work even if there're other workers inserting new repos.
* workflows/ci: Valgrind tests should pass nowEgor Tensin2023-07-06
|
* cmake: fix Valgrind runs with Clang buildsEgor Tensin2023-07-06
|
* test: try mitigating port clashesEgor Tensin2023-07-06
| | | | Also, I don't think calling random.seed is necessary.