Commits

Author Commit Message Labels Comments Date
Armin Rigo
Add the (now-passing) test from issue #92
Armin Rigo
Found and fixed the bug.
Armin Rigo
I think that this should clean up the (relative) mess of the GC on some-but-not-all cdata objects. The test still doesn't pass though.
Armin Rigo
A test, and start to work on a fix, which doesn't work so far
Armin Rigo
Mention Win64.
Armin Rigo
Tweak the error message
Armin Rigo
Finish the fix for issue #91.
Armin Rigo
merge heads
Armin Rigo
Document ffi.new_handle().
Armin Rigo
Fix the front-end to work (but still show $ signs)
Armin Rigo
Attack issue 91; first step.
David Schneider
extend and fix test_c.py for ARM
David Schneider
check for ARM using a flag like it is done for MSVC
David Schneider
Fix for the alignment calculation of anonymous and zero sized bitfields on ARM.
Armin Rigo
Propagate the original OSError, which contains information returned by dlerror().
Armin Rigo
Fix MSVC bitfields in all tested cases.
Armin Rigo
Fix this test for MSVC
Armin Rigo
MSVC-style bitfields, first attempt
Armin Rigo
Always run the C compiler, even if it's MSVC. Found out that the logic to pack bitfields is, of course, very different.
Armin Rigo
One more test
Armin Rigo
Oups, fix.
Armin Rigo
issue #82: implement bitfields in a way that is, as far as I can tell, fully compatible with gcc.
Armin Rigo
Fixes to the test
Armin Rigo
Improve the tests
Armin Rigo
Failing tests for issue #82.
Armin Rigo
Issue #81 resolved: dir(ffi.verify(...))
Armin Rigo
issue 87: first stab
Armin Rigo
Bump the version number of 0.7.
Armin Rigo
Fix (motivated by a pypy without cpyext): if we're using vengine_gen.py, then we should not use imp.find_module() to locate our compiled module, because it's actually not a CPython C API module at all. Instead, just walk sys.path in that case.
Armin Rigo
Test and fix: two pointers of different types can compare equal, but they still had different hashes
  1. Prev
  2. Next