Fortran: procedures with BIND(C) attribute require explicit interface [PR85877]

  Dear all,

the Fortran standard requires an explicit procedure interface in certain
situations, such as when they have a BIND(C) attribute (F2018:15.4.2.2).
The attached patch adds a check for this.

Regtested on x86_64-pc-linux-gnu.  OK for mainline?

The PR marks this as a long-time regression, so it might be backported
to all open branches.

Thanks,
Harald
  

Hi Harald,

> the Fortran standard requires an explicit procedure interface in certain
> situations, such as when they have a BIND(C) attribute (F2018:15.4.2.2).
> The attached patch adds a check for this.
> 
> Regtested on x86_64-pc-linux-gnu.  OK for mainline?

While this fixes the ICE, it misses

function f() bind(c)
   f = 42.
end

subroutine p
   bind(c) f     ! { dg-error "must be explicit" }
   x = f()
end

subroutine s
   interface
      function g() bind(c)
      end function g
   end interface
   x = g()
end

where the interface is picked up via a global symbol.  This code
may not be valid; nagfor rejects it, but I cannot find a
constraint at least in the F2022 draft that prohibits it.

Hm... might it be better to check for attr->module_procedure ||
attr->if_source == IFSRC_IFBODY?

Best regards

	Thomas
  

Hi Thomas,

Am 18.03.23 um 19:52 schrieb Thomas Koenig via Gcc-patches:
> Hi Harald,
>
>> the Fortran standard requires an explicit procedure interface in certain
>> situations, such as when they have a BIND(C) attribute (F2018:15.4.2.2).
>> The attached patch adds a check for this.
>>
>> Regtested on x86_64-pc-linux-gnu.  OK for mainline?
>
> While this fixes the ICE, it misses
>
> function f() bind(c)
>    f = 42.
> end
>
> subroutine p
>    bind(c) f     ! { dg-error "must be explicit" }
>    x = f()
> end

what do you mean by "it misses"?  I do not see an explicit interface
here, only a global symbol.  All compiler I tried with the above
code reject it one way or the other, e g. Cray:

   x = f()
   ^
ftn-954 crayftn: ERROR P, File = pr85877-2.f90, Line = 12, Column = 3
   Procedure "f", referenced at line 6 (pr85877-2.f90) must have an
explicit interface because one or more arguments have the BIND attribute.


> subroutine s
>    interface
>       function g() bind(c)
>       end function g
>    end interface
>    x = g()
> end
>
> where the interface is picked up via a global symbol.

Is it really true that this is in the spirit of the standard?
Is the global declaration above really equivalent to an explicit
interface?

I would expect legal code to be like:

function g() bind(c)
   g = 42.
end

subroutine s
   interface
      function g() bind(c)
      end function g
   end interface
   x = g()
end

unless you do it in the context of a module and in the right way.

> This code
> may not be valid; nagfor rejects it, but I cannot find a
> constraint at least in the F2022 draft that prohibits it.

I thought that F2018:15.4.2.2 and F2023:15.4.2.2(7) are rather
clear and "explicit" on this.  IMHO the case of "f" above seems
to be excluded and thus not conforming.

> Hm... might it be better to check for attr->module_procedure ||
> attr->if_source == IFSRC_IFBODY?

Maybe we should find a set of legal and illegal cases that we
can agree upon.

Thanks,
Harald

> Best regards
>
>      Thomas
>
>
  

Hi Harald,

> Am 18.03.23 um 19:52 schrieb Thomas Koenig via Gcc-patches:
>> Hi Harald,
>>
>>> the Fortran standard requires an explicit procedure interface in certain
>>> situations, such as when they have a BIND(C) attribute (F2018:15.4.2.2).
>>> The attached patch adds a check for this.
>>>
>>> Regtested on x86_64-pc-linux-gnu.  OK for mainline?
>>
>> While this fixes the ICE, it misses
>>
>> function f() bind(c)
>>    f = 42.
>> end
>>
>> subroutine p
>>    bind(c) f     ! { dg-error "must be explicit" }
>>    x = f()
>> end
> 
> what do you mean by "it misses"?

Sorry, that was caused by confusion on my part (and it is better
to test an assumption of what the compiler actually does :-)

Patch is OK, also for backport.  Maybe you can also include the
test above, just to make sure.

Best regards

	Thomas
  

Hi Thomas,

Am 19.03.23 um 08:34 schrieb Thomas Koenig via Gcc-patches:
> Hi Harald,
>
>> Am 18.03.23 um 19:52 schrieb Thomas Koenig via Gcc-patches:
>>> Hi Harald,
>>>
>>>> the Fortran standard requires an explicit procedure interface in
>>>> certain
>>>> situations, such as when they have a BIND(C) attribute
>>>> (F2018:15.4.2.2).
>>>> The attached patch adds a check for this.
>>>>
>>>> Regtested on x86_64-pc-linux-gnu.  OK for mainline?
>>>
>>> While this fixes the ICE, it misses
>>>
>>> function f() bind(c)
>>>    f = 42.
>>> end
>>>
>>> subroutine p
>>>    bind(c) f     ! { dg-error "must be explicit" }
>>>    x = f()
>>> end
>>
>> what do you mean by "it misses"?
>
> Sorry, that was caused by confusion on my part (and it is better
> to test an assumption of what the compiler actually does :-)
>
> Patch is OK, also for backport.  Maybe you can also include the
> test above, just to make sure.

I've added your suggestion to the testcase.  Pushed as:

https://gcc.gnu.org/g:5426ab34643d9e6502f3ee572891a03471fa33ed

> Best regards
>
>      Thomas
>
>

Thanks,
Harald