PR69455

[dhumieres.dominique@free.fr]

Le 2019-10-18 15:56, Steve Kargl a écrit :
> On Fri, Oct 18, 2019 at 01:01:18PM +0200, dhumieres.dominique@free.fr 
> wrote:
>> The patch fixes the problem for iso_c_binding but not for
>> ISO_FORTRAN_ENV (pr69455#c1).
>> 
>> The later is fixed with
>> 
>> --- ../_clean/gcc/fortran/trans-decl.c	2019-10-11 11:28:20.000000000
>> +0200
>> +++ gcc/fortran/trans-decl.c	2019-10-18 12:52:46.000000000 +0200
>> @@ -5987,7 +5987,12 @@ generate_local_decl (gfc_symbol * sym)
>> 
>>         if (sym->ns && sym->ns->construct_entities)
>>   	{
>> -	  if (sym->attr.referenced)
>> +	  /* Construction of the ISO_C_BINDING module within a BLOCK
>> +	     construct where ONLY and RENAMED entities are included in
>> +	     seems to be bogus.  See PR 69455.  */
>> +	  if (sym->attr.referenced
>> +	      && sym->from_intmod != INTMOD_ISO_C_BINDING
>> +	      && sym->from_intmod != INTMOD_ISO_FORTRAN_ENV)
>>   	    gfc_get_symbol_decl (sym);
>>   	  sym->mark = 1;
>>   	}
>> 
>> Note that I did not updated the comment nor the test.
>> 
> 
> If you don't have test code, then how do you know that your
> modification fixes anything?  Without a testcase, I'll be
> committing my original patch.

The test case is in pr69455 comment 1:

block
   use, intrinsic :: ISO_FORTRAN_ENV, only: wp => REAL32, ik => INT32
   print *, ik, wp
end block

block
   use, intrinsic :: ISO_FORTRAN_ENV, only: wp => REAL64, ik => INT64
   print *, ik, wp
end block
end

Dominique