Recent clf ICE report

[Steve Kargl]

On Thu, Dec 16, 2010 at 06:55:21AM +0100, Paul Richard Thomas wrote:
> Steve,
> 
> It is worth recording the response of the Intel compiler to you reduced version:
> 
> [pault@localhost tmp]# ifort -static j*.f90
> james1.f90(4): error #7300: This derived-type object in an
> input-output list cannot have private components.   [TRANSFER]
>   print *, transfer(32512, C_NULL_PTR)
> -----------^
> compilation aborted for james1.f90 (code 1)
> 

Thanks, Paul.

I had not thought about possible IO issues.  One still gets
an ICE with

program bug1
   use ISO_C_BINDING
   implicit none
   type(c_ptr) :: m
   m = transfer(32512, m)
end program bug1 

James has posted a few more examples in the clf thread.  One
he seems to think is standard conforming.  I'm still trying to 
understand the standard to determine whether I agree with James.

-- 
Steve