Bug 6772 – Cannot pass cfloat argument type to a function on x86_64

In the example below the function gets a different value for argument n than value passed cf. Argument type and size is OK, but value is wrong. Both D1 and D2 have this problem. Problem only present for 64 bit code. import std.stdio; void myfunc(cfloat n) { writefln("argument type is %s", typeof(n).stringof); writefln("n = 0x%X, sizeof = %d", *(cast(ulong*)&n), n.sizeof); writefln("real part %f, imag part %f", n.re, n.im); } void main() { cfloat cf; cf = 1.02f+3.04fi; writefln("cf = 0x%X, sizeof = %d", *(cast(ulong*)&cf), cf.sizeof); myfunc(cf); }

This is working with the latest.