summaryrefslogtreecommitdiff
path: root/testing
Commit message (Collapse)AuthorAgeFilesLines
* Issue 556Armin Rigo2023-01-091-4/+2
| | | | Trying to make this test a little bit less fragile
* Issue 553Armin Rigo2022-12-041-0/+18
| | | | Two missing calls to PyObject_GC_UnTrack()
* Drop py.code usage from tests, no longer depend on the deprecated py packageMiro Hrončok2022-11-142-3/+4
|
* Replace py.test usage with pytest, explicitly require py for testspy.testMiro Hrončok2022-11-1125-590/+595
| | | | | | | pytest 7.2+ no longer depends on py. It ships py.path and py.error only. See https://docs.pytest.org/en/7.2.x/changelog.html#deprecations The tests use py.code as well, hence we declare and document a test dependency on py.
* Tentative fix for issue #542Armin Rigo2022-07-181-2/+11
|
* musllinux test updates and wheels, py3.11 beta wheelsMatt Davis2022-06-299-13/+46
|
* Test and fix for the case where ffi.embedding_api() is called butArmin Rigo2021-12-283-1/+22
| | | | does not list any `extern "Python"` function
* fix tuple version and tests to support rcMatt Davis2021-09-211-0/+1
|
* fix setuptools tests for macossetuptools_test_macos_fixMatt Davis2021-09-031-1/+4
|
* #507Armin Rigo2021-09-031-0/+2
| | | | Test failure on ia64
* fix tests for py.test running on CPython v3.10.0b3Armin Rigo2021-06-202-7/+8
|
* another test fixArmin Rigo2021-06-201-3/+4
|
* source-only update, to version number 1.14.4-1v1.14.4-1Armin Rigo2020-11-291-1/+6
|
* add missing includememsetMatti Picus2020-11-011-0/+1
|
* Issue #478Armin Rigo2020-10-291-0/+5
| | | | Fix a case of nested anonymous struct/unions when recompiling to Python
* Revert 58f2e09a4254 and part of 4a15fb9e4aec: it was fixed in CPython after theArmin Rigo2020-07-283-14/+0
| | | | 3.9b5 release.
* more CPython 3.9b5 fixesArmin Rigo2020-07-272-2/+7
|
* CPython 3.9b5 fixArmin Rigo2020-07-272-0/+9
|
* forgot this fileArmin Rigo2020-07-241-0/+26
|
* pff I love Python 3Armin Rigo2020-07-241-1/+2
|
* Allow ffi.embedding_init_code("unicode-with-non-ascii-chars")Armin Rigo2020-07-241-0/+6
|
* more of the same, for win64Armin Rigo2020-07-211-4/+9
|
* fix this logicArmin Rigo2020-07-211-4/+1
|
* Skip this test by default. It keeps finding corner cases that are bugs ofArmin Rigo2020-07-211-6/+9
| | | | libffi but not cffi
* pff, it seems that "running for more than 0.2 seconds" is also a failure ↵Armin Rigo2020-07-211-2/+2
| | | | | | | condition, which hypothesis then tries to reduce, giving not reasonable answer because here one test run of any size takes around 0.2 seconds
* A new set of segfaults on Windows using callbacksArmin Rigo2020-07-211-2/+37
|
* Keep sys.path, otherwise it's testing another version of _cffi_backend.soArmin Rigo2020-07-191-0/+1
|
* Test the callbacks, tooArmin Rigo2020-07-191-2/+41
|
* force calls via libffi. shows the problem recently solved on windowsArmin Rigo2020-07-191-4/+7
|
* Check with the real C compilerArmin Rigo2020-07-192-89/+124
|
* Add a hypothesis test. Mostly checks libffi, or our own libffi_msvc on windowsArmin Rigo2020-07-191-0/+89
|
* ctypes on windows doesn't correctly return 3-bytes structsArmin Rigo2020-07-191-0/+2
|
* Extra tests for using a 3-bytes struct as a return typeArmin Rigo2020-07-182-0/+47
|
* on darwin, arm64 bitfields act like x86_64Lawrence D'Anna2020-07-091-2/+12
|
* #454Armin Rigo2020-05-291-0/+41
| | | | Second try with '# NUMBER' instead of '#line NUMBER', as gcc seems to output
* Issue #454Armin Rigo2020-05-261-1/+47
| | | | Try harder to avoid #line directives confuse the rest of pre-parsing
* #453Armin Rigo2020-05-201-0/+34
| | | | | | Special-case typedefs of arrays with '...' length, so that they can be used through recompiler.py as if they had a specified length in most cases.
* copy this from pypyArmin Rigo2020-04-161-0/+3
|
* Merge branch 'branch/msvcrt' into 'branch/default'Armin Rigo2020-04-161-3/+3
|\ | | | | | | | | | | Branch/msvcrt See merge request pypy/cffi!100
| * add missing importmsvcrtMatti Picus2020-04-011-1/+1
| |
| * support python2 builds with msvc>13Matti Picus2020-04-011-2/+2
| |
* | Vendor in pypy's version of py.path.local.make_numbered_dir(), which is moreArmin Rigo2020-03-061-2/+129
|/ | | | resilient against random OS Errors like EACCES
* May fix test on OS/XArmin Rigo2020-02-071-0/+1
|
* Issue #440Armin Rigo2020-01-193-0/+33
| | | | | | Limit the amount of memory that is requested from alloca() for temporary conversion of arguments. Non-small requests are instead handled with PyObject_Malloc() and PyObject_Free().
* Python 3.9 compatArmin Rigo2020-01-074-13/+22
|
* merge branch 'shortername', thanks mattiArmin Rigo2020-01-061-1/+1
|\
| * tweak how to pass define_macros, shorten module name to not go over 260 charsMatti Picus2020-01-051-1/+1
| |
* | fix some gcc warningsArmin Rigo2020-01-051-8/+11
| |
* | fix test for b9a98b040bd7Armin Rigo2020-01-051-3/+3
| |
* | Issue #437Armin Rigo2020-01-052-0/+47
|/ | | | Support ffi.dlopen(<void* cdata>). See updated documentation.
View Lorry Controller Status