Fix a latent bug in the asm code. I think that technically, using "*m0" instead of "*0" allows LLVM to pick between using the same memory as output 0 and using a new memory location. (So far I haven't been able to construct a testcase that actually breaks because of this, though)

module mini.norun_debug5;

void main()
{
    int i = 32;
    real r = 3.1415;
    real* p = &r;
    func(i,r,p);
}

void func(int i, real r, real* p)
{
    int* fail;
    *fail = 666;
}