pybind11

Commit Graph

Author	SHA1	Message	Date
Henry Schreiner	c50f90eca6	style: use Black everywhere (#2594 ) * style: use Black everywhere * style: minor touchup from review	2020-10-16 16:38:13 -04:00
Henry Schreiner	deba040b6f	test: hide segfault (#2559 ) * tests: Don't run tests that often segfault * tests: drop all cross module gil tests * tests: try skipping all macOS Python 3.9 tests * tests: drop macOS Python 3.9	2020-10-08 12:25:04 -04:00
Henry Fredrick Schreiner	e3774b76ed	tests: the condition was too look on xfail	2020-09-16 23:21:47 -04:00
Henry Schreiner	3a89bffac0	ci: harden chrono test, mark another macos 4.9 dev failure (#2448 ) * ci: harden chrono test, mark another macos 4.9 dev failure This should help with a little of the flakiness seen with the timing test * Update tests/test_chrono.py * Can also fail	2020-08-31 14:28:07 -04:00
Henry Schreiner	110e6c12ce	ci: reduce flakiness a little (#2418 )	2020-08-20 11:58:34 -04:00
Henry Schreiner	d8c7ee00a6	ci: GHA basic format & pre-commit (#2309 )	2020-07-20 13:35:21 -04:00
Saran Tunyasuvunakool	b60fd233fa	Make sure `detail::get_internals` acquires the GIL before making Python calls. (#1836 ) This is only necessary if `get_internals` is called for the first time in a given module when the running thread is in a GIL-released state. Fixes #1364	2019-07-15 16:47:02 +02:00
Yannick Jadoul	085a29436a	Increasing timeout in test_gil_scoped.py to get AppVeyor to succeed	2019-01-03 22:43:52 +01:00
Borja Zarco	e2b884c33b	Use `PyGILState_GetThisThreadState` when using gil_scoped_acquire. (#1211 ) This avoids GIL deadlocking when pybind11 tries to acquire the GIL in a thread that already acquired it using standard Python API (e.g. when running from a Python thread).	2018-12-01 22:47:40 +09:00

9 Commits