1. 23 Nov, 2022 1 commit
    • Ralf W. Grosse-Kunstleve's avatar
      Use `multiprocessing` `start_method` `"forkserver"` (#4306) · 9c18a74e
      Ralf W. Grosse-Kunstleve authored
      * Use `multiprocessing` `start_method` `"forkserver"`
      
      Alternative to PR #4305
      
      * Add link to comment under PR #4105
      
      * Unconditionally `pytest.skip("DEADLOCK")` for PyPy Windows
      
      * Remove `SKIP_IF_DEADLOCK` entirely, for simplicity. Hopefully this PR will resolve the deadlocks for good.
      
      * Add "In a nutshell" comment, in response to request by @EricCousineau-TRI
      9c18a74e
  2. 22 Nov, 2022 1 commit
  3. 12 Nov, 2022 2 commits
  4. 10 Nov, 2022 1 commit
  5. 02 Nov, 2022 1 commit
    • Ethan Steinberg's avatar
      Fix functional.h bug + introduce test to verify that it is fixed (#4254) · ee2b5226
      Ethan Steinberg authored
      
      
      * Illustrate bug in functional.h
      
      * style: pre-commit fixes
      
      * Make functional casting more robust / add workaround
      
      * Make function_record* casting even more robust
      
      * See if this fixes PyPy issue
      
      * It still fails on PyPy sadly
      
      * Do not make new CTOR just yet
      
      * Fix test
      
      * Add name to ensure correctness
      
      * style: pre-commit fixes
      
      * Clean up tests + remove ifdef guards
      
      * Add comments
      
      * Improve comments, error handling, and safety
      
      * Fix compile error
      
      * Fix magic logic
      
      * Extract helper function
      
      * Fix func signature
      
      * move to local internals
      
      * style: pre-commit fixes
      
      * Switch to simpler design
      
      * style: pre-commit fixes
      
      * Move to function_record
      
      * style: pre-commit fixes
      
      * Switch to internals, update tests and docs
      
      * Fix lint
      
      * Oops, forgot to resolve last comment
      
      * Fix typo
      
      * Update in response to comments
      
      * Implement suggestion to improve test
      
      * Update comment
      
      * Simple fixes
      Co-authored-by: default avatarpre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
      Co-authored-by: default avatarAaron Gokaslan <aaronGokaslan@gmail.com>
      ee2b5226
  6. 01 Nov, 2022 2 commits
  7. 31 Oct, 2022 3 commits
  8. 30 Oct, 2022 2 commits
    • Ralf W. Grosse-Kunstleve's avatar
    • Chekov2k's avatar
      Add `PYBIND11_SIMPLE_GIL_MANAGEMENT` option (cmake, C++ define) (#4216) · b07d08f6
      Chekov2k authored
      * Add option to force the use of the PYPY GIL scoped acquire/release logic to support nested gil access, see https://github.com/pybind/pybind11/issues/1276 and https://github.com/pytorch/pytorch/issues/83101
      
      
      
      * Apply suggestions from code review
      
      * Update CMakeLists.txt
      
      * docs: update upgrade guide
      
      * Update docs/upgrade.rst
      
      * All bells & whistles.
      
      * Add Reminder to common.h, so that we will not forget to purge `!WITH_THREAD` branches when dropping Python 3.6
      
      * New sentence instead of semicolon.
      
      * Temporarily pull in snapshot of PR #4246
      
      * Add `test_release_acquire`
      
      * Add more unit tests for nested gil locking
      
      * Add test_report_builtins_internals_keys
      
      * Very minor enhancement: sort list only after filtering.
      
      * Revert change in docs/upgrade.rst
      
      * Add test_multi_acquire_release_cross_module, while also forcing unique PYBIND11_INTERNALS_VERSION for cross_module_gil_utils.cpp
      
      * Hopefully fix apparently new ICC error.
      
      ```
      2022-10-28T07:57:54.5187728Z -- The CXX compiler identification is Intel 2021.7.0.20220726
      ...
      2022-10-28T07:58:53.6758994Z icpc: remark #10441: The Intel(R) C++ Compiler Classic (ICC) is deprecated and will be removed from product release in the second half of 2023. The Intel(R) oneAPI DPC++/C++ Compiler (ICX) is the recommended compiler moving forward. Please transition to use this compiler. Use '-diag-disable=10441' to disable this message.
      2022-10-28T07:58:54.5801597Z In file included from /home/runner/work/pybind11/pybind11/include/pybind11/detail/../detail/type_caster_base.h(15),
      2022-10-28T07:58:54.5803794Z                  from /home/runner/work/pybind11/pybind11/include/pybind11/detail/../cast.h(15),
      2022-10-28T07:58:54.5805740Z                  from /home/runner/work/pybind11/pybind11/include/pybind11/detail/../attr.h(14),
      2022-10-28T07:58:54.5809556Z                  from /home/runner/work/pybind11/pybind11/include/pybind11/detail/class.h(12),
      2022-10-28T07:58:54.5812154Z                  from /home/runner/work/pybind11/pybind11/include/pybind11/pybind11.h(13),
      2022-10-28T07:58:54.5948523Z                  from /home/runner/work/pybind11/pybind11/tests/cross_module_gil_utils.cpp(13):
      2022-10-28T07:58:54.5949009Z /home/runner/work/pybind11/pybind11/include/pybind11/detail/../detail/internals.h(177): error #2282: unrecognized GCC pragma
      2022-10-28T07:58:54.5949374Z       PYBIND11_TLS_KEY_INIT(tstate)
      2022-10-28T07:58:54.5949579Z       ^
      2022-10-28T07:58:54.5949695Z
      ```
      
      * clang-tidy fixes
      
      * Workaround for PYPY WIN exitcode None
      
      * Revert "Temporarily pull in snapshot of PR #4246"
      
      This reverts commit 23ac16e859150f27fda25ca865cabcb4444e0770.
      
      * Another workaround for PYPY WIN exitcode None
      
      * Clean up how the tests are run "run in process" Part 1: uniformity
      
      * Clean up how the tests are run "run in process" Part 2: use `@pytest.mark.parametrize` and clean up the naming.
      
      * Skip some tests `#if defined(THREAD_SANITIZER)` (tested with TSAN using the Google-internal toolchain).
      
      * Run all tests again but ignore ThreadSanitizer exitcode 66 (this is less likely to mask unrelated ThreadSanitizer issues in the future).
      
      * bug fix: missing common.h include before using `PYBIND11_SIMPLE_GIL_MANAGEMENT`
      
      For the tests in the github CI this does not matter, because
      `PYBIND11_SIMPLE_GIL_MANAGEMENT` is always defined from the command line,
      but when monkey-patching common.h locally, it matters.
      
      * if process.exitcode is None: assert t_delta > 9.9
      
      * More sophisiticated `_run_in_process()` implementation, clearly reporting `DEADLOCK`, additionally exercised via added `intentional_deadlock()`
      
      * Wrap m.intentional_deadlock in a Python function, for `ForkingPickler` compatibility.
      
      ```
      >       ForkingPickler(file, protocol).dump(obj)
      E       TypeError: cannot pickle 'PyCapsule' object
      ```
      
      Observed with all Windows builds including mingw but not PyPy, and macos-latest with Python 3.9, 3.10, 3.11 but not 3.6.
      
      * Add link to potential solution for WOULD-BE-NICE-TO-HAVE feature.
      
      * Add `SKIP_IF_DEADLOCK = True` option, to not pollute the CI results with expected `DEADLOCK` failures while we figure out what to do about them.
      
      * Add COPY-PASTE-THIS: gdb ... command (to be used for debugging the detected deadlock)
      
      * style: pre-commit fixes
      
      * Do better than automatic pre-commit fixes.
      
      * Add `PYBIND11_SIMPLE_GIL_MANAGEMENT` to `pytest_report_header()` (so that we can easily know when harvesting deadlock information from the CI logs).
      Co-authored-by: default avatarArnim Balzer <arnim@seechange.ai>
      Co-authored-by: default avatarHenry Schreiner <HenrySchreinerIII@gmail.com>
      Co-authored-by: default avatarRalf W. Grosse-Kunstleve <rwgk@google.com>
      Co-authored-by: default avatarpre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
      b07d08f6
  9. 29 Oct, 2022 1 commit
  10. 26 Oct, 2022 1 commit
  11. 23 Oct, 2022 3 commits
  12. 22 Oct, 2022 1 commit
  13. 21 Oct, 2022 3 commits
  14. 20 Oct, 2022 5 commits
  15. 18 Oct, 2022 1 commit
    • Lalaland's avatar
      First draft of Eigen::Tensor support (#4201) · fab1eebe
      Lalaland authored
      * First draft of Eigen::Tensor support
      
      * [pre-commit.ci] auto fixes from pre-commit.com hooks
      
      for more information, see https://pre-commit.ci
      
      * Fix build errors
      
      * Weird allocator stuff?
      
      * [pre-commit.ci] auto fixes from pre-commit.com hooks
      
      for more information, see https://pre-commit.ci
      
      * Remove unused + additional allocator junk
      
      * Disable warning
      
      * Use constexpr
      
      * [pre-commit.ci] auto fixes from pre-commit.com hooks
      
      for more information, see https://pre-commit.ci
      
      * clang tidy fixes
      
      * [pre-commit.ci] auto fixes from pre-commit.com hooks
      
      for more information, see https://pre-commit.ci
      
      * Resolve comments
      
      * [pre-commit.ci] auto fixes from pre-commit.com hooks
      
      for more information, see https://pre-commit.ci
      
      * Remove auto constexpr function
      
      * [pre-commit.ci] auto fixes from pre-commit.com hooks
      
      for more information, see https://pre-commit.ci
      
      * Try again for older C++
      
      * Oops forgot constexpr
      
      * Move...
      fab1eebe
  16. 17 Oct, 2022 2 commits
  17. 12 Oct, 2022 2 commits
    • Ralf W. Grosse-Kunstleve's avatar
      Minor `py::capsule` cleanup. No functional change. (#4238) · 964c4997
      Ralf W. Grosse-Kunstleve authored
      Use `PyCapsule_Destructor` (part of the stable Python ABI) instead of spelling out the C `typedef`.
      
      The deprecation message is misleading. Replace with a message pointing to another existing ctor.
      
      Background: According to @wjakob the original motivation for deprecating the ctor (in PR #752) was to hide Python C API details, but PR #902 brought those back with a new ctor, it cannot be avoided. Having a `PyCapsule_Destructor` or a `void (*destructor)(void *)` are two separate and valid use cases.
      964c4997
    • Aaron Gokaslan's avatar
      8781daf6
  18. 11 Oct, 2022 2 commits
  19. 10 Oct, 2022 1 commit
    • Ralf W. Grosse-Kunstleve's avatar
      Reproducer and fix for issue encountered in smart_holder update. (#4228) · da104a9e
      Ralf W. Grosse-Kunstleve authored
      * Reproducer for issue encountered in smart_holder update.
      
      * clang-tidy compatibility (untested).
      
      * Add `enable_if_t` to workaround.
      
      * Bug fix: Move `PYBIND11_USING_WORKAROUND_FOR_CUDA_11_4_THROUGH_8` determination to detail/common.h
      
      So that it actually is defined in pybind11.h
      
      * Try using the workaround (which is nicer than the original code) universally.
      
      * Reduce reproducer for CUDA 11.7 issue encountered in smart_holder update.
      
      This commit tested in isolation on top of current master + first version of reproducer (62311eb431849d135a5db84f6c75ec390f2ede7c).
      
      Succeeds with Debian Clang 14.0.6 C++17 (and probably all other compilers).
      
      Fails for CUDA 11.7:
      
      ```
      cd /build/tests && /usr/local/cuda/bin/nvcc -forward-unknown-to-host-compiler -Dpybind11_tests_EXPORTS -I/mounted_pybind11/include -isystem=/usr/include/python3.10 -g --generate-code=arch=compute_52,code=[compute_52,sm_52] -Xcompiler=-fPIC -Xcompiler=-fvisibility=hidden -Werror all-warnings -std=c++17 -MD -MT tests/CMakeFiles/pybind11_tests.dir/test_class.cpp.o -MF CMakeFiles/pybind11_tests.dir/test_class.cpp.o.d -x cu -c /mounted_pybind11/tests/test_class.cpp -o CMakeFiles/pybind11_tests.dir/test_class.cpp.o
      /mounted_pybind11/tests/test_class.cpp(53): error: more than one instance of overloaded function "pybind11::class_<type_, options...>::def [with type_=test_class::pr4220_tripped_over_this::Empty0, options=<>]" matches the argument list:
                  function template "pybind11::class_<test_class::pr4220_tripped_over_this::Empty0> &pybind11::class_<type_, options...>::def(const char *, Func &&, const Extra &...) [with type_=test_class::pr4220_tripped_over_this::Empty0, options=<>]"
      /mounted_pybind11/include/pybind11/pybind11.h(1557): here
                  function template "pybind11::class_<test_class::pr4220_tripped_over_this::Empty0> &pybind11::class_<type_, options...>::def(const T &, const Extra &...) [with type_=test_class::pr4220_tripped_over_this::Empty0, options=<>]"
      /mounted_pybind11/include/pybind11/pybind11.h(1586): here
                  argument types are: (const char [8], <unknown-type>)
                  object type is: pybind11::class_<test_class::pr4220_tripped_over_this::Empty0>
      
      1 error detected in the compilation of "/mounted_pybind11/tests/test_class.cpp".
      ```
      da104a9e
  20. 07 Oct, 2022 3 commits
  21. 06 Oct, 2022 1 commit
  22. 04 Oct, 2022 1 commit