Michael: When posting to the bug tracker, please trim irrelevant parts of old messages. It makes it hard to see if you actually added anything new.

> [Me] If your compiler does not support “signed short” bitfields, maybe we just have to accept that ctypes supports it even though the compiler doesn’t, and skip the test.

Proper support for “signed short” according to standard C I guess would mean if you define

struct BITS {
    signed short M: 1;
} b;
b.M = -1;

then reading back b.M gives -1. However I realized test_bitfields tests overflowing values rather than negative values. In any case, I think we have established that neither of these cases work with XLC.

I haven’t changed the signed int A–I fields yet. That was part of my patch. I was waiting for confirmation about the __xlC__ check, before committing the whole thing.

Eric: I proposed to conditionally skip the test; see disable-signed-short.patch. Since many other compilers apparently pass the test and support signed short, we should probably keep the test.