summaryrefslogtreecommitdiff
path: root/testing/test_verify.py
Commit message (Collapse)AuthorAgeFilesLines
* fix for win32Matti Picus2014-03-021-4/+9
|
* Skip test that crashes the interpreter on sparcStefano Rivera2014-01-221-0/+3
|
* An extra testArmin Rigo2014-01-141-0/+21
|
* merge headsArmin Rigo2014-01-071-4/+4
|\
| * Link math tests using `sin()` to `libm`Vincent Bernat2013-12-211-4/+4
| | | | | | | | | | | | | | | | | | While this linking seems to be done indirectly on most platforms, it fails to work correctly on Sparc and S390x where an inappropriate version of `sin()` is used. This should additional occurrences of bug #68. Tests are passing on amd64, sparc and s390x Debian Linux platforms. Both Python 2.x and Python 3.x.
* | Kill is_{signed,unsigned}_type()Armin Rigo2013-12-151-17/+17
|/
* "alloca.h" does not exist on FreeBSD. Thanks Tobias.Armin Rigo2013-11-171-1/+5
|
* Test various combinations of calls. Mostly a libffi stress-test.Armin Rigo2013-11-121-0/+84
|
* Skip half the test with MSVCArmin Rigo2013-11-121-0/+2
|
* Ah bahArmin Rigo2013-11-121-1/+1
|
* Issue #118: improve the detection and error message, jumpingArmin Rigo2013-11-101-0/+13
| | | | through hoops to cover both signed and unsigned cases.
* Test and fix: a remaining deadlockArmin Rigo2013-11-091-2/+5
|
* Fix test.Armin Rigo2013-11-081-1/+1
|
* Fix vengine_gen for global arrays with [...].Armin Rigo2013-11-081-1/+6
|
* Fix some testsArmin Rigo2013-11-081-8/+8
|
* Some target testsArmin Rigo2013-11-061-11/+50
|
* Issue #113: Test and fix: Values of anonymous enums are not exposedArmin Rigo2013-10-261-0/+15
|
* Support partial unions in a way very similar to partial structs.Armin Rigo2013-10-171-0/+30
| | | | Needed for a python-cffi mail.
* (arigo, fijal, jerith) More complicated (and better) juggling of compiler ↵Jeremy Thurgood2013-10-051-7/+10
| | | | args for gcc/clang in verify tests.
* Two tests from the mailing list, and fixes: bogus "const" detection,Armin Rigo2013-09-141-0/+13
| | | | and global array variables should never be "const".
* issue 102: allow ffi.typeof() to work on functions taking a structArmin Rigo2013-09-111-0/+10
| | | | | argument (with vengine_gen). Note that it still fails on functions taking an auto-completed struct. Unsure if and how to fix.
* Finish the fix for issue #91.Armin Rigo2013-06-191-1/+1
|
* Issue #81 resolved: dir(ffi.verify(...))Armin Rigo2013-05-301-0/+15
|
* Fix for enums used as bitfields. (thanks jerith)Armin Rigo2013-05-051-0/+12
|
* py3 compatPhilip Jenvey2013-04-051-1/+1
|
* Python 3 compatArmin Rigo2013-04-011-1/+3
|
* Merge d0f6755466f7 into default:Armin Rigo2013-03-311-2/+25
|\ | | | | | | | | | | | | | | | | Cancel again these changes: we can't pass 0 or None to mean NULL. Just use NULL explicitly. (After discussion on IRC, notably with Amaury -- thanks) Removed some unrelated doc changes that describe something that was planned but never implemented.
| * Skip testing for large unsigned values on Windows, where the C compilerArmin Rigo2013-03-301-0/+2
| | | | | | | | silently truncates them to negative values, apparently.
| * Only ignore mismatches between *different* types; using twice the same type ↵Armin Rigo2013-03-301-1/+1
| | | | | | | | must be tested to work.
| * Ignore mismatches between two different primitive types of the same size,Armin Rigo2013-03-301-1/+6
| | | | | | | | like on some platforms "double" and "long double".
| * Be more lenient and skip this test in non-x86 casesArmin Rigo2013-03-301-1/+2
| |
| * A passing test.Armin Rigo2013-03-281-0/+6
| |
| * A bit hackish, but solves exactly issue #71: ffi.typeof(builtin_function).Armin Rigo2013-03-281-0/+8
| |
* | Cancel again these changes: we can't pass 0 or None to mean NULL. JustArmin Rigo2013-03-071-5/+5
|/ | | | | use NULL explicitly. (After discussion on IRC, notably with Amaury -- thanks)
* Fix the test for 64-bit: alignment issuesArmin Rigo2013-03-031-4/+5
|
* Generate correct type name for pointers to arrays.Tom Nixon2013-03-021-0/+8
| | | | | | | | | Previously, get_c_name("a") on the type "int (*)[5]" would return "int *[5]". This caused "incompatible type" errors when verifying structs containing pointers to arrays. A partial workaround for this bug was originally made in vengine_*; this has been removed.
* Change the hack: instead of passing None for NULL pointers, we pass 0.Armin Rigo2013-02-281-3/+9
| | | | | This is more compatible with C. Also accept 0 at other places that expect a pointer.
* Python 3 fixes.Armin Rigo2013-02-281-1/+1
|
* Moving the determination of the base integer type of an enumArmin Rigo2013-02-161-0/+29
| | | | | | out of C code, into the 'cffi' package. This is work in progress; it should eventually permit the new test to fully pass. For now wrote a warning in the doc.
* hg merge auto-typesArmin Rigo2013-02-121-14/+23
|\
| * ImplementationArmin Rigo2013-02-081-14/+23
| |
* | An extra test for 'bool'.auto-typesArmin Rigo2013-02-121-0/+1
|/
* Passing None as an ``item *`` argument to a function.Armin Rigo2013-02-081-0/+16
|
* * Win32 fixArmin Rigo2013-02-071-1/+8
| | | | * Don't depend on the order in which the two calls are made!
* Python 3 compatArmin Rigo2013-02-071-3/+3
|
* Extra testsArmin Rigo2013-01-071-0/+4
|
* Test and fix for issue #51: unlike what is documented, we can't pass aArmin Rigo2013-01-071-0/+20
| | | | | list argument to a function expecting a "foo *" argument. It doesn't work on CPython (but works with vengine_gen or on PyPy).
* issue #46: don't pass -Werror on recent MacsArmin Rigo2013-01-011-3/+10
|
* Try harder to run the tests even if PYTHONPATH was not set and cffi notArmin Rigo2013-01-011-2/+3
| | | | | installed. Tested with a "python setup.py build_ext -i" as well as with a fresh "python setup.py install".
* Document how to indirectly define callbacks using unsupportedArmin Rigo2012-12-141-0/+36
| | | | features
View Lorry Controller Status