Commit message (Collapse) | Author | Age | Files | Lines | |
---|---|---|---|---|---|
* | another attempt at Py_LIMITED_API on Windows on CPython >= 3.5 | Armin Rigo | 2020-08-13 | 1 | -12/+41 |
| | |||||
* | Issue #440 | Armin Rigo | 2020-01-19 | 1 | -0/+48 |
| | | | | | | 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(). | ||||
* | Tweak the '-Wno-*' arguments passed to gcc during tests | Armin Rigo | 2019-11-07 | 1 | -2/+2 |
| | |||||
* | On Windows, we still can't use Py_LIMITED_API for bad reasons. | Armin Rigo | 2019-02-16 | 1 | -35/+12 |
| | |||||
* | Redo 164e526a5515 and 14ce6985e1c3. | Armin Rigo | 2018-08-24 | 1 | -3/+36 |
| | |||||
* | [windows] Support Py_LIMITED_API | Ofek Lev | 2018-08-19 | 1 | -10/+0 |
| | |||||
* | Don't define py_limited_api at all on Windows, for now. | Armin Rigo | 2018-01-13 | 1 | -1/+2 |
| | |||||
* | Issue #355 | Armin Rigo | 2018-01-13 | 1 | -29/+11 |
| | | | | Revert 164e526a5515 and 14ce6985e1c3 and add an explanation about why. | ||||
* | Only do that dance if _MSC_VER is defined | Armin Rigo | 2018-01-08 | 1 | -5/+5 |
| | |||||
* | Issue #350: more mess on Windows with Py_LIMITED_API | Armin Rigo | 2018-01-08 | 1 | -3/+30 |
| | |||||
* | Issue #337 | Armin Rigo | 2017-09-30 | 1 | -3/+3 |
| | | | | Fixes for test failures and warnings on ARM | ||||
* | Issue #300 | Armin Rigo | 2017-07-21 | 1 | -0/+1 |
| | | | | | Hopefully fix the remaining cases where a _Bool return value was not correctly converted to a Python bool, but still gave a Python int. | ||||
* | Didn't figure out how to cleanly define charN_t inside _cffi_include.h. | Armin Rigo | 2017-06-04 | 1 | -5/+0 |
| | | | | So for now, don't. | ||||
* | Fix: the type char32_t is defined by the standard as always unsigned | Armin Rigo | 2017-06-02 | 1 | -1/+1 |
| | |||||
* | tweaks | Armin Rigo | 2017-05-31 | 1 | -8/+8 |
| | |||||
* | in-progress | Armin Rigo | 2017-05-31 | 1 | -1/+50 |
| | |||||
* | (dstufft) Don't overwrite a user-specified value for Py_LIMITED_API, if | Armin Rigo | 2017-04-11 | 1 | -1/+1 |
| | | | | any | ||||
* | Don't pollute the namespace with 'CTypeDescrObject' | Armin Rigo | 2016-11-05 | 1 | -9/+10 |
| | |||||
* | If we say Py_LIMITED_API and we're compiling with a debug version of | Armin Rigo | 2016-09-03 | 1 | -0/+16 |
| | | | | CPython, "#include <Python.h>" crashes. Work around it the hard way. | ||||
* | Avoid a combination where we end up with two incompatible "typedef xxx | Armin Rigo | 2016-08-06 | 1 | -1/+3 |
| | | | | _Bool;" | ||||
* | Seems that as a #define, this hack breaks if a future header does | Armin Rigo | 2016-08-06 | 1 | -1/+1 |
| | | | | | "typedef bool _Bool;". With a typedef, though, we get two identical typedefs, which at least g++ is perfectly happy with | ||||
* | Use Py_LIMITED_API in the generated C extension modules, because doing | Armin Rigo | 2016-07-19 | 1 | -14/+0 |
| | | | | | so seems easy. Still need to check if and how we have to produce the actual .so files. | ||||
* | Test and fix: the "bool" type in C++ | Armin Rigo | 2016-06-05 | 1 | -0/+6 |
| | |||||
* | Issue #245: __stdcall not generated correctly on extern "Python" functions | Armin Rigo | 2016-02-04 | 1 | -0/+6 |
| | |||||
* | in-progress | Armin Rigo | 2015-12-02 | 1 | -1/+2 |
| | |||||
* | First version | Armin Rigo | 2015-12-01 | 1 | -1/+2 |
| | |||||
* | update the code to use 'extern "Python"' | Armin Rigo | 2015-11-18 | 1 | -2/+2 |
| | |||||
* | tweak the pypy-only generated code | Armin Rigo | 2015-11-15 | 1 | -1/+2 |
| | |||||
* | tweak tweak tweak until we can at least read the function pointer out of | Armin Rigo | 2015-11-13 | 1 | -2/+6 |
| | | | | the 'lib' object | ||||
* | Test the various cases, fix issues with rounding, etc. | Armin Rigo | 2015-08-28 | 1 | -3/+4 |
| | | | | | | A subtle point is that your version uses is_float_type(), which is true also for "long double", and so it will loose precision on platforms like x86 Linux where "long double" is larger than "double". | ||||
* | Add support for `typedef float... foo_t`. | Max Belanger | 2015-08-27 | 1 | -0/+5 |
| | |||||
* | Issue #216: add "defined(__hpux)" for alloca.h | Armin Rigo | 2015-08-13 | 1 | -1/+1 |
| | |||||
* | Tweaks and minor fixes | Armin Rigo | 2015-05-30 | 1 | -1/+1 |
| | |||||
* | Issues 69, 73: add the syntax "typedef int... foo_t;". | Armin Rigo | 2015-05-30 | 1 | -6/+4 |
| | |||||
* | Performance: no real need to call PyArg_ParseTuple() here | Armin Rigo | 2015-05-26 | 1 | -6/+19 |
| | |||||
* | Silence some warnings with gcc 5.1: if T is a type smaller than 'int', | Armin Rigo | 2015-05-26 | 1 | -2/+3 |
| | | | | | | | then the C expression "condition ? (T)x : (T)y" is typed not as a T, but as a larger 'int' --- which means there was an implicit cast from the final 'int' to the T-typed target of the assignment whenever this macro was used. | ||||
* | Fix comments | Armin Rigo | 2015-05-22 | 1 | -1/+2 |
| | |||||
* | Defines the macro _CFFI_ before including Python.h. | Armin Rigo | 2015-05-22 | 1 | -0/+1 |
| | |||||
* | Best-effort attempt at supporting C++. There is still one issue shown | Armin Rigo | 2015-05-17 | 1 | -2/+9 |
| | | | | in test_recompiler if we replace "if 0:" with "if 1:". | ||||
* | the big Moving Files Around step | Armin Rigo | 2015-05-12 | 1 | -0/+202 |