Hi Steve,
I see that the implementation of the standard is slightly more
complicated than I thought.
type :: t(a,b)
integer, kind :: a
integer, len :: b
integer(a) :: v(b)
end type t
type(t(4,:)), allocatable :: z1
type(t(4,10)), allocatable :: z2
allocate (t(4, :) :: z1) ! { dg-error "cannot contain DEFERRED parameters" }
allocate (t(4, *) :: z2) ! This should give an error because it isn't a dummy.
call foo (z1, z2, z2)
contains
subroutine foo (arg1, arg2, arg3)
type(t(4,:)), allocatable :: arg1
type(t(4,*)), allocatable :: arg2
type(t(4,10)), allocatable :: arg3
allocate (t(4, :) :: arg1) ! { dg-error "cannot contain DEFERRED
parameters" }
allocate (t(4, *) :: arg2) ! This needs to be handled correctly at
runtime but is legal.
allocate (t(4, *) :: arg2) ! This should give an error because the
dummy parameter is not assumed.
end subroutine
end
I'll hit this next weekend because I am away all this week.
Regards
Paul
On 9 December 2017 at 19:55, Steve Kargl
<s...@troutmask.apl.washington.edu> wrote:
> On Fri, Dec 08, 2017 at 05:13:28PM -0800, Steve Kargl wrote:
>> The attached patch enforces F2008:C631, which of course is
>>
>> /* F2008:C631 (R626) A type-param-value in a type-spec shall be an
>> asterisk if and only if each allocate-object is a dummy argument
>> for which the corresponding type parameter is assumed. */
>>
>> Regression tested on x86_64-*-freebsd.
>>
>> 2017-12-08 Steven G. Kargl <ka...@gcc.gnu.org>
>>
>> PR fortran/82934
>> PR fortran/83318
>> * match.c (gfc_match_allocate): Enforce F2008:C631.
>>
>> 2017-12-08 Steven G. Kargl <ka...@gcc.gnu.org>
>>
>> PR fortran/82934
>> PR fortran/83318
>> * gfortran.dg/allocate_assumed_charlen_2.f90: new test.
>>
>
> The final version of the patch that I committed is attached.
>
> --
> Steve
--
"If you can't explain it simply, you don't understand it well enough"
- Albert Einstein
