[patch, fortran] PR41859 ICE on invalid expression involving DT with pointer components in I/O

[Jerry DeLisle]

On 08/17/2010 10:58 PM, Paul Richard Thomas wrote:
> Dear Jerry,
>
>> OK for trunk?
>
> OK but for one tiny question:
>
> Should you not check for parentheses, just to stop the code trying to
> charge down, for example, binary operators?  Since expr1 has been
> resolved already at the top of resolve_code, this should not matter -
> it's just unnecessary.
>

I was thinking of a case like this:

   print *, (-p)

Where p is invalid.  We need to drill down to run the check on it.

I also found I get a regression with integer_exponentiation.f90 if we do not 
drill down. This turns out to have a substatial mix of operators.  Interesting 
test case.

Thanks for review.

Jerry