[Haskell-cafe] Re: FFI and C99 "complex"

[Marcin 'Qrczak' Kowalczyk]

Aaron Denney <wnoise at ofb.net> writes:

> So, I have heard claims that C 99 specifies that a variable v
> of type "complex t" is stored as if declared as t v[2], with
> v[0] the real part and v[1] the imaginary part.
>
> I don't have a copy of the spec at hand.  Could someone who does verify
> this?

This is true (in C99 draft; I don't think this has changed in the
final standard).

-- 
   __("<         Marcin Kowalczyk
   \__/       qrczak at knm.org.pl
    ^^     http://qrnik.knm.org.pl/~qrczak/